Merge branch 'wip-v0.4' into cc/0.4/docs

Lots of work that wasn't directly related to core
improvements/messages/testing functionality

.devcontainer/README.md

@@ -15,12 +15,12 @@ You may use the button above, or follow these steps to open this repo in a Codes
 1. Click **Create codespace on master**.
 
 For more info, check out the [GitHub documentation](https://docs.github.com/en/free-pro-team@latest/github/developing-online-with-codespaces/creating-a-codespace#creating-a-codespace).
-  
+
 ## VS Code Dev Containers
 
 [![Open in Dev Containers](https://img.shields.io/static/v1?label=Dev%20Containers&message=Open&color=blue&logo=visualstudiocode)](https://vscode.dev/redirect?url=vscode://ms-vscode-remote.remote-containers/cloneInVolume?url=https://github.com/langchain-ai/langchain)
 
-> [!NOTE]  
+> [!NOTE]
 > If you click the link above you will open the main repo (`langchain-ai/langchain`) and *not* your local cloned repo. This is fine if you only want to run and test the library, but if you want to contribute you can use the link below and replace with your username and cloned repo name:
 
 ```txt

.devcontainer/docker-compose.yaml

@@ -4,7 +4,7 @@ services:
     build:
       dockerfile: libs/langchain/dev.Dockerfile
       context: ..
-    
+
     networks:
       - langchain-network
 

.github/CODE_OF_CONDUCT.md

@@ -129,4 +129,4 @@ For answers to common questions about this code of conduct, see the FAQ at
 [v2.1]: https://www.contributor-covenant.org/version/2/1/code_of_conduct.html
 [Mozilla CoC]: https://github.com/mozilla/diversity
 [FAQ]: https://www.contributor-covenant.org/faq
-[translations]: https://www.contributor-covenant.org/translations
+[translations]: https://www.contributor-covenant.org/translations

.github/ISSUE_TEMPLATE/bug-report.yml

@@ -5,7 +5,7 @@ body:
   - type: markdown
     attributes:
       value: |
-        Thank you for taking the time to file a bug report. 
+        Thank you for taking the time to file a bug report.
 
         Use this to report BUGS in LangChain. For usage questions, feature requests and general design questions, please use the [LangChain Forum](https://forum.langchain.com/).
 
@@ -50,7 +50,7 @@ body:
 
         If a maintainer can copy it, run it, and see it right away, there's a much higher chance that you'll be able to get help.
 
-        **Important!** 
+        **Important!**
 
         * Avoid screenshots when possible, as they are hard to read and (more importantly) don't allow others to copy-and-paste your code.
         * Reduce your code to the minimum required to reproduce the issue if possible. This makes it much easier for others to help you.
@@ -58,14 +58,14 @@ body:
         * INCLUDE the language label (e.g. `python`) after the first three backticks to enable syntax highlighting. (e.g., ```python rather than ```).
 
       placeholder: |
-        The following code: 
+        The following code:
 
         ```python
         from langchain_core.runnables import RunnableLambda
 
         def bad_code(inputs) -> int:
           raise NotImplementedError('For demo purpose')
-          
+
           chain = RunnableLambda(bad_code)
           chain.invoke('Hello!')
         ```

.github/ISSUE_TEMPLATE/documentation.yml

@@ -14,7 +14,7 @@ body:
 
         Do **NOT** use this to ask usage questions or reporting issues with your code.
 
-        If you have usage questions or need help solving some problem, 
+        If you have usage questions or need help solving some problem,
         please use the [LangChain Forum](https://forum.langchain.com/).
 
         If you're in the wrong place, here are some helpful links to find a better

.github/ISSUE_TEMPLATE/privileged.yml

@@ -8,7 +8,7 @@ body:
 
         If you are not a LangChain maintainer or were not asked directly by a maintainer to create an issue, then please start the conversation on the [LangChain Forum](https://forum.langchain.com/) instead.
 
-        You are a LangChain maintainer if you maintain any of the packages inside of the LangChain repository 
+        You are a LangChain maintainer if you maintain any of the packages inside of the LangChain repository
         or are a regular contributor to LangChain with previous merged pull requests.
   - type: checkboxes
     id: privileged

.github/actions/people/Dockerfile

@@ -4,4 +4,4 @@ RUN pip install httpx PyGithub "pydantic==2.0.2" pydantic-settings "pyyaml>=5.3.
 
 COPY ./app /app
 
-CMD ["python", "/app/main.py"]
+CMD ["python", "/app/main.py"]

.github/actions/people/action.yml

@@ -4,8 +4,8 @@ description: "Generate the data for the LangChain People page"
 author: "Jacob Lee <jacob@langchain.dev>"
 inputs:
   token:
-    description: 'User token, to read the GitHub API. Can be passed in using {{ secrets.LANGCHAIN_PEOPLE_GITHUB_TOKEN }}'
+    description: "User token, to read the GitHub API. Can be passed in using {{ secrets.LANGCHAIN_PEOPLE_GITHUB_TOKEN }}"
     required: true
 runs:
-  using: 'docker'
-  image: 'Dockerfile'
+  using: "docker"
+  image: "Dockerfile"

.github/scripts/check_diff.py

@@ -3,14 +3,12 @@ import json
 import os
 import sys
 from collections import defaultdict
-from typing import Dict, List, Set
 from pathlib import Path
+from typing import Dict, List, Set
+
 import tomllib
-
-from packaging.requirements import Requirement
-
 from get_min_versions import get_min_version_from_toml
-
+from packaging.requirements import Requirement
 
 LANGCHAIN_DIRS = [
     "libs/core",
@@ -38,7 +36,7 @@ IGNORED_PARTNERS = [
 ]
 
 PY_312_MAX_PACKAGES = [
-    "libs/partners/chroma", # https://github.com/chroma-core/chroma/issues/4382
+    "libs/partners/chroma",  # https://github.com/chroma-core/chroma/issues/4382
 ]
 
 
@@ -85,9 +83,9 @@ def dependents_graph() -> dict:
                 for depline in extended_deps:
                     if depline.startswith("-e "):
                         # editable dependency
-                        assert depline.startswith(
-                            "-e ../partners/"
-                        ), "Extended test deps should only editable install partner packages"
+                        assert depline.startswith("-e ../partners/"), (
+                            "Extended test deps should only editable install partner packages"
+                        )
                         partner = depline.split("partners/")[1]
                         dep = f"langchain-{partner}"
                     else:
@@ -271,7 +269,7 @@ if __name__ == "__main__":
                     dirs_to_run["extended-test"].add(dir_)
         elif file.startswith("libs/standard-tests"):
             # TODO: update to include all packages that rely on standard-tests (all partner packages)
-            # note: won't run on external repo partners
+            # Note: won't run on external repo partners
             dirs_to_run["lint"].add("libs/standard-tests")
             dirs_to_run["test"].add("libs/standard-tests")
             dirs_to_run["lint"].add("libs/cli")
@@ -285,7 +283,7 @@ if __name__ == "__main__":
         elif file.startswith("libs/cli"):
             dirs_to_run["lint"].add("libs/cli")
             dirs_to_run["test"].add("libs/cli")
-            
+
         elif file.startswith("libs/partners"):
             partner_dir = file.split("/")[2]
             if os.path.isdir(f"libs/partners/{partner_dir}") and [
@@ -303,7 +301,10 @@ if __name__ == "__main__":
                 f"Unknown lib: {file}. check_diff.py likely needs "
                 "an update for this new library!"
             )
-        elif file.startswith("docs/") or file in ["pyproject.toml", "uv.lock"]: # docs or root uv files
+        elif file.startswith("docs/") or file in [
+            "pyproject.toml",
+            "uv.lock",
+        ]:  # docs or root uv files
             docs_edited = True
             dirs_to_run["lint"].add(".")
 

.github/scripts/check_prerelease_dependencies.py

@@ -1,4 +1,5 @@
 import sys
+
 import tomllib
 
 if __name__ == "__main__":

.github/scripts/get_min_versions.py

@@ -1,5 +1,5 @@
-from collections import defaultdict
 import sys
+from collections import defaultdict
 from typing import Optional
 
 if sys.version_info >= (3, 11):
@@ -8,17 +8,13 @@ else:
     # for python 3.10 and below, which doesnt have stdlib tomllib
     import tomli as tomllib
 
-from packaging.requirements import Requirement
-from packaging.specifiers import SpecifierSet
-from packaging.version import Version
-
-
-import requests
-from packaging.version import parse
+import re
 from typing import List
 
-import re
-
+import requests
+from packaging.requirements import Requirement
+from packaging.specifiers import SpecifierSet
+from packaging.version import Version, parse
 
 MIN_VERSION_LIBS = [
     "langchain-core",
@@ -72,11 +68,13 @@ def get_minimum_version(package_name: str, spec_string: str) -> Optional[str]:
     spec_string = re.sub(r"\^0\.0\.(\d+)", r"0.0.\1", spec_string)
     # rewrite occurrences of ^0.y.z to >=0.y.z,<0.y+1 (can be anywhere in constraint string)
     for y in range(1, 10):
-        spec_string = re.sub(rf"\^0\.{y}\.(\d+)", rf">=0.{y}.\1,<0.{y+1}", spec_string)
+        spec_string = re.sub(
+            rf"\^0\.{y}\.(\d+)", rf">=0.{y}.\1,<0.{y + 1}", spec_string
+        )
     # rewrite occurrences of ^x.y.z to >=x.y.z,<x+1.0.0 (can be anywhere in constraint string)
     for x in range(1, 10):
         spec_string = re.sub(
-            rf"\^{x}\.(\d+)\.(\d+)", rf">={x}.\1.\2,<{x+1}", spec_string
+            rf"\^{x}\.(\d+)\.(\d+)", rf">={x}.\1.\2,<{x + 1}", spec_string
         )
 
     spec_set = SpecifierSet(spec_string)
@@ -169,12 +167,12 @@ def check_python_version(version_string, constraint_string):
     # rewrite occurrences of ^0.y.z to >=0.y.z,<0.y+1.0 (can be anywhere in constraint string)
     for y in range(1, 10):
         constraint_string = re.sub(
-            rf"\^0\.{y}\.(\d+)", rf">=0.{y}.\1,<0.{y+1}.0", constraint_string
+            rf"\^0\.{y}\.(\d+)", rf">=0.{y}.\1,<0.{y + 1}.0", constraint_string
         )
     # rewrite occurrences of ^x.y.z to >=x.y.z,<x+1.0.0 (can be anywhere in constraint string)
     for x in range(1, 10):
         constraint_string = re.sub(
-            rf"\^{x}\.0\.(\d+)", rf">={x}.0.\1,<{x+1}.0.0", constraint_string
+            rf"\^{x}\.0\.(\d+)", rf">={x}.0.\1,<{x + 1}.0.0", constraint_string
         )
 
     try:

.github/scripts/prep_api_docs_build.py

@@ -3,9 +3,10 @@
 
 import os
 import shutil
-import yaml
 from pathlib import Path
-from typing import Dict, Any
+from typing import Any, Dict
+
+import yaml
 
 
 def load_packages_yaml() -> Dict[str, Any]:
@@ -28,7 +29,6 @@ def get_target_dir(package_name: str) -> Path:
 def clean_target_directories(packages: list) -> None:
     """Remove old directories that will be replaced."""
     for package in packages:
-
         target_dir = get_target_dir(package["name"])
         if target_dir.exists():
             print(f"Removing {target_dir}")
@@ -38,7 +38,6 @@ def clean_target_directories(packages: list) -> None:
 def move_libraries(packages: list) -> None:
     """Move libraries from their source locations to the target directories."""
     for package in packages:
-
         repo_name = package["repo"].split("/")[1]
         source_path = package["path"]
         target_dir = get_target_dir(package["name"])
@@ -68,23 +67,33 @@ def main():
         package_yaml = load_packages_yaml()
 
         # Clean target directories
-        clean_target_directories([
-            p
-            for p in package_yaml["packages"]
-            if (p["repo"].startswith("langchain-ai/") or p.get("include_in_api_ref"))
-            and p["repo"] != "langchain-ai/langchain"
-            and p["name"] != "langchain-ai21"  # Skip AI21 due to dependency conflicts
-        ])
+        clean_target_directories(
+            [
+                p
+                for p in package_yaml["packages"]
+                if (
+                    p["repo"].startswith("langchain-ai/") or p.get("include_in_api_ref")
+                )
+                and p["repo"] != "langchain-ai/langchain"
+                and p["name"]
+                != "langchain-ai21"  # Skip AI21 due to dependency conflicts
+            ]
+        )
 
         # Move libraries to their new locations
-        move_libraries([
-            p
-            for p in package_yaml["packages"]
-            if not p.get("disabled", False)
-            and (p["repo"].startswith("langchain-ai/") or p.get("include_in_api_ref"))
-            and p["repo"] != "langchain-ai/langchain"
-            and p["name"] != "langchain-ai21"  # Skip AI21 due to dependency conflicts
-        ])
+        move_libraries(
+            [
+                p
+                for p in package_yaml["packages"]
+                if not p.get("disabled", False)
+                and (
+                    p["repo"].startswith("langchain-ai/") or p.get("include_in_api_ref")
+                )
+                and p["repo"] != "langchain-ai/langchain"
+                and p["name"]
+                != "langchain-ai21"  # Skip AI21 due to dependency conflicts
+            ]
+        )
 
         # Delete ones without a pyproject.toml
         for partner in Path("langchain/libs/partners").iterdir():

.github/tools/git-restore-mtime

@@ -81,56 +81,93 @@ import time
 __version__ = "2022.12+dev"
 
 # Update symlinks only if the platform supports not following them
-UPDATE_SYMLINKS = bool(os.utime in getattr(os, 'supports_follow_symlinks', []))
+UPDATE_SYMLINKS = bool(os.utime in getattr(os, "supports_follow_symlinks", []))
 
 # Call os.path.normpath() only if not in a POSIX platform (Windows)
-NORMALIZE_PATHS = (os.path.sep != '/')
+NORMALIZE_PATHS = os.path.sep != "/"
 
 # How many files to process in each batch when re-trying merge commits
 STEPMISSING = 100
 
 # (Extra) keywords for the os.utime() call performed by touch()
-UTIME_KWS = {} if not UPDATE_SYMLINKS else {'follow_symlinks': False}
+UTIME_KWS = {} if not UPDATE_SYMLINKS else {"follow_symlinks": False}
 
 
 # Command-line interface ######################################################
 
+
 def parse_args():
-    parser = argparse.ArgumentParser(
-        description=__doc__.split('\n---')[0])
+    parser = argparse.ArgumentParser(description=__doc__.split("\n---")[0])
 
     group = parser.add_mutually_exclusive_group()
-    group.add_argument('--quiet', '-q', dest='loglevel',
-        action="store_const", const=logging.WARNING, default=logging.INFO,
-        help="Suppress informative messages and summary statistics.")
-    group.add_argument('--verbose', '-v', action="count", help="""
+    group.add_argument(
+        "--quiet",
+        "-q",
+        dest="loglevel",
+        action="store_const",
+        const=logging.WARNING,
+        default=logging.INFO,
+        help="Suppress informative messages and summary statistics.",
+    )
+    group.add_argument(
+        "--verbose",
+        "-v",
+        action="count",
+        help="""
         Print additional information for each processed file.
         Specify twice to further increase verbosity.
-        """)
+        """,
+    )
 
-    parser.add_argument('--cwd', '-C', metavar="DIRECTORY", help="""
+    parser.add_argument(
+        "--cwd",
+        "-C",
+        metavar="DIRECTORY",
+        help="""
         Run as if %(prog)s was started in directory %(metavar)s.
         This affects how --work-tree, --git-dir and PATHSPEC arguments are handled.
         See 'man 1 git' or 'git --help' for more information.
-        """)
+        """,
+    )
 
-    parser.add_argument('--git-dir', dest='gitdir', metavar="GITDIR", help="""
+    parser.add_argument(
+        "--git-dir",
+        dest="gitdir",
+        metavar="GITDIR",
+        help="""
         Path to the git repository, by default auto-discovered by searching
         the current directory and its parents for a .git/ subdirectory.
-        """)
+        """,
+    )
 
-    parser.add_argument('--work-tree', dest='workdir', metavar="WORKTREE", help="""
+    parser.add_argument(
+        "--work-tree",
+        dest="workdir",
+        metavar="WORKTREE",
+        help="""
         Path to the work tree root, by default the parent of GITDIR if it's
         automatically discovered, or the current directory if GITDIR is set.
-        """)
+        """,
+    )
 
-    parser.add_argument('--force', '-f', default=False, action="store_true", help="""
+    parser.add_argument(
+        "--force",
+        "-f",
+        default=False,
+        action="store_true",
+        help="""
         Force updating files with uncommitted modifications.
         Untracked files and uncommitted deletions, renames and additions are
         always ignored.
-        """)
+        """,
+    )
 
-    parser.add_argument('--merge', '-m', default=False, action="store_true", help="""
+    parser.add_argument(
+        "--merge",
+        "-m",
+        default=False,
+        action="store_true",
+        help="""
         Include merge commits.
         Leads to more recent times and more files per commit, thus with the same
         time, which may or may not be what you want.
@@ -138,71 +175,130 @@ def parse_args():
         are found sooner, which can improve performance, sometimes substantially.
         But as merge commits are usually huge, processing them may also take longer.
         By default, merge commits are only used for files missing from regular commits.
-        """)
+        """,
+    )
 
-    parser.add_argument('--first-parent', default=False, action="store_true", help="""
+    parser.add_argument(
+        "--first-parent",
+        default=False,
+        action="store_true",
+        help="""
         Consider only the first parent, the "main branch", when evaluating merge commits.
         Only effective when merge commits are processed, either when --merge is
         used or when finding missing files after the first regular log search.
         See --skip-missing.
-        """)
+        """,
+    )
 
-    parser.add_argument('--skip-missing', '-s', dest="missing", default=True,
-        action="store_false", help="""
+    parser.add_argument(
+        "--skip-missing",
+        "-s",
+        dest="missing",
+        default=True,
+        action="store_false",
+        help="""
         Do not try to find missing files.
         If merge commits were not evaluated with --merge and some files were
         not found in regular commits, by default %(prog)s searches for these
         files again in the merge commits.
         This option disables this retry, so files found only in merge commits
         will not have their timestamp updated.
-        """)
+        """,
+    )
 
-    parser.add_argument('--no-directories', '-D', dest='dirs', default=True,
-        action="store_false", help="""
+    parser.add_argument(
+        "--no-directories",
+        "-D",
+        dest="dirs",
+        default=True,
+        action="store_false",
+        help="""
         Do not update directory timestamps.
         By default, use the time of its most recently created, renamed or deleted file.
         Note that just modifying a file will NOT update its directory time.
-        """)
+        """,
+    )
 
-    parser.add_argument('--test', '-t', default=False, action="store_true",
-        help="Test run: do not actually update any file timestamp.")
+    parser.add_argument(
+        "--test",
+        "-t",
+        default=False,
+        action="store_true",
+        help="Test run: do not actually update any file timestamp.",
+    )
 
-    parser.add_argument('--commit-time', '-c', dest='commit_time', default=False,
-        action='store_true', help="Use commit time instead of author time.")
+    parser.add_argument(
+        "--commit-time",
+        "-c",
+        dest="commit_time",
+        default=False,
+        action="store_true",
+        help="Use commit time instead of author time.",
+    )
 
-    parser.add_argument('--oldest-time', '-o', dest='reverse_order', default=False,
-        action='store_true', help="""
+    parser.add_argument(
+        "--oldest-time",
+        "-o",
+        dest="reverse_order",
+        default=False,
+        action="store_true",
+        help="""
         Update times based on the oldest, instead of the most recent commit of a file.
         This reverses the order in which the git log is processed to emulate a
         file "creation" date. Note this will be inaccurate for files deleted and
         re-created at later dates.
-        """)
+        """,
+    )
 
-    parser.add_argument('--skip-older-than', metavar='SECONDS', type=int, help="""
+    parser.add_argument(
+        "--skip-older-than",
+        metavar="SECONDS",
+        type=int,
+        help="""
         Ignore files that are currently older than %(metavar)s.
         Useful in workflows that assume such files already have a correct timestamp,
         as it may improve performance by processing fewer files.
-        """)
+        """,
+    )
 
-    parser.add_argument('--skip-older-than-commit', '-N', default=False,
-        action='store_true', help="""
+    parser.add_argument(
+        "--skip-older-than-commit",
+        "-N",
+        default=False,
+        action="store_true",
+        help="""
         Ignore files older than the timestamp it would be updated to.
         Such files may be considered "original", likely in the author's repository.
-        """)
+        """,
+    )
 
-    parser.add_argument('--unique-times', default=False, action="store_true", help="""
+    parser.add_argument(
+        "--unique-times",
+        default=False,
+        action="store_true",
+        help="""
         Set the microseconds to a unique value per commit.
         Allows telling apart changes that would otherwise have identical timestamps,
         as git's time accuracy is in seconds.
-        """)
+        """,
+    )
 
-    parser.add_argument('pathspec', nargs='*', metavar='PATHSPEC', help="""
+    parser.add_argument(
+        "pathspec",
+        nargs="*",
+        metavar="PATHSPEC",
+        help="""
         Only modify paths matching %(metavar)s, relative to current directory.
         By default, update all but untracked files and submodules.
-        """)
+        """,
+    )
 
-    parser.add_argument('--version', '-V', action='version',
-        version='%(prog)s version {version}'.format(version=get_version()))
+    parser.add_argument(
+        "--version",
+        "-V",
+        action="version",
+        version="%(prog)s version {version}".format(version=get_version()),
+    )
 
     args_ = parser.parse_args()
     if args_.verbose:
@@ -212,17 +308,18 @@ def parse_args():
 
 
 def get_version(version=__version__):
-    if not version.endswith('+dev'):
+    if not version.endswith("+dev"):
         return version
     try:
         cwd = os.path.dirname(os.path.realpath(__file__))
-        return Git(cwd=cwd, errors=False).describe().lstrip('v')
+        return Git(cwd=cwd, errors=False).describe().lstrip("v")
     except Git.Error:
-        return '-'.join((version, "unknown"))
+        return "-".join((version, "unknown"))
 
 
 # Helper functions ############################################################
 
+
 def setup_logging():
     """Add TRACE logging level and corresponding method, return the root logger"""
     logging.TRACE = TRACE = logging.DEBUG // 2
@@ -255,11 +352,13 @@ def normalize(path):
     if path and path[0] == '"':
         # Python 2: path = path[1:-1].decode("string-escape")
         # Python 3: https://stackoverflow.com/a/46650050/624066
-        path = (path[1:-1]                 # Remove enclosing double quotes
-                .encode('latin1')          # Convert to bytes, required by 'unicode-escape'
-                .decode('unicode-escape')  # Perform the actual octal-escaping decode
-                .encode('latin1')          # 1:1 mapping to bytes, UTF-8 encoded
-                .decode('utf8', 'surrogateescape'))  # Decode from UTF-8
+        path = (
+            path[1:-1]  # Remove enclosing double quotes
+            .encode("latin1")  # Convert to bytes, required by 'unicode-escape'
+            .decode("unicode-escape")  # Perform the actual octal-escaping decode
+            .encode("latin1")  # 1:1 mapping to bytes, UTF-8 encoded
+            .decode("utf8", "surrogateescape")
+        )  # Decode from UTF-8
     if NORMALIZE_PATHS:
         # Make sure the slash matches the OS; for Windows we need a backslash
         path = os.path.normpath(path)
@@ -282,12 +381,12 @@ def touch_ns(path, mtime_ns):
 
 def isodate(secs: int):
     # time.localtime() accepts floats, but discards fractional part
-    return time.strftime('%Y-%m-%d %H:%M:%S', time.localtime(secs))
+    return time.strftime("%Y-%m-%d %H:%M:%S", time.localtime(secs))
 
 
 def isodate_ns(ns: int):
     # for integers fromtimestamp() is equivalent and ~16% slower than isodate()
-    return datetime.datetime.fromtimestamp(ns / 1000000000).isoformat(sep=' ')
+    return datetime.datetime.fromtimestamp(ns / 1000000000).isoformat(sep=" ")
 
 
 def get_mtime_ns(secs: int, idx: int):
@@ -305,35 +404,49 @@ def get_mtime_path(path):
 
 # Git class and parse_log(), the heart of the script ##########################
 
+
 class Git:
     def __init__(self, workdir=None, gitdir=None, cwd=None, errors=True):
-        self.gitcmd = ['git']
+        self.gitcmd = ["git"]
         self.errors = errors
         self._proc = None
-        if workdir: self.gitcmd.extend(('--work-tree', workdir))
-        if gitdir:  self.gitcmd.extend(('--git-dir',   gitdir))
-        if cwd:     self.gitcmd.extend(('-C',          cwd))
+        if workdir:
+            self.gitcmd.extend(("--work-tree", workdir))
+        if gitdir:
+            self.gitcmd.extend(("--git-dir", gitdir))
+        if cwd:
+            self.gitcmd.extend(("-C", cwd))
         self.workdir, self.gitdir = self._get_repo_dirs()
 
     def ls_files(self, paths: list = None):
-        return (normalize(_) for _ in self._run('ls-files --full-name', paths))
+        return (normalize(_) for _ in self._run("ls-files --full-name", paths))
 
     def ls_dirty(self, force=False):
-        return (normalize(_[3:].split(' -> ', 1)[-1])
-                for _ in self._run('status --porcelain')
-                if _[:2] != '??' and (not force or (_[0] in ('R', 'A')
-                                                    or _[1] == 'D')))
+        return (
+            normalize(_[3:].split(" -> ", 1)[-1])
+            for _ in self._run("status --porcelain")
+            if _[:2] != "??" and (not force or (_[0] in ("R", "A") or _[1] == "D"))
+        )
 
-    def log(self, merge=False, first_parent=False, commit_time=False,
-            reverse_order=False, paths: list = None):
-        cmd = 'whatchanged --pretty={}'.format('%ct' if commit_time else '%at')
-        if merge:         cmd += ' -m'
-        if first_parent:  cmd += ' --first-parent'
-        if reverse_order: cmd += ' --reverse'
+    def log(
+        self,
+        merge=False,
+        first_parent=False,
+        commit_time=False,
+        reverse_order=False,
+        paths: list = None,
+    ):
+        cmd = "whatchanged --pretty={}".format("%ct" if commit_time else "%at")
+        if merge:
+            cmd += " -m"
+        if first_parent:
+            cmd += " --first-parent"
+        if reverse_order:
+            cmd += " --reverse"
         return self._run(cmd, paths)
 
     def describe(self):
-        return self._run('describe --tags', check=True)[0]
+        return self._run("describe --tags", check=True)[0]
 
     def terminate(self):
         if self._proc is None:
@@ -345,18 +458,22 @@ class Git:
             pass
 
     def _get_repo_dirs(self):
-        return (os.path.normpath(_) for _ in
-            self._run('rev-parse --show-toplevel --absolute-git-dir', check=True))
+        return (
+            os.path.normpath(_)
+            for _ in self._run(
+                "rev-parse --show-toplevel --absolute-git-dir", check=True
+            )
+        )
 
     def _run(self, cmdstr: str, paths: list = None, output=True, check=False):
         cmdlist = self.gitcmd + shlex.split(cmdstr)
         if paths:
-            cmdlist.append('--')
+            cmdlist.append("--")
             cmdlist.extend(paths)
-        popen_args = dict(universal_newlines=True, encoding='utf8')
+        popen_args = dict(universal_newlines=True, encoding="utf8")
         if not self.errors:
-            popen_args['stderr'] = subprocess.DEVNULL
-        log.trace("Executing: %s", ' '.join(cmdlist))
+            popen_args["stderr"] = subprocess.DEVNULL
+        log.trace("Executing: %s", " ".join(cmdlist))
         if not output:
             return subprocess.call(cmdlist, **popen_args)
         if check:
@@ -379,30 +496,26 @@ def parse_log(filelist, dirlist, stats, git, merge=False, filterlist=None):
     mtime = 0
     datestr = isodate(0)
     for line in git.log(
-            merge,
-            args.first_parent,
-            args.commit_time,
-            args.reverse_order,
-            filterlist
+        merge, args.first_parent, args.commit_time, args.reverse_order, filterlist
     ):
-        stats['loglines'] += 1
+        stats["loglines"] += 1
 
         # Blank line between Date and list of files
         if not line:
             continue
 
         # Date line
-        if line[0] != ':':  # Faster than `not line.startswith(':')`
-            stats['commits'] += 1
+        if line[0] != ":":  # Faster than `not line.startswith(':')`
+            stats["commits"] += 1
             mtime = int(line)
             if args.unique_times:
-                mtime = get_mtime_ns(mtime, stats['commits'])
+                mtime = get_mtime_ns(mtime, stats["commits"])
             if args.debug:
                 datestr = isodate(mtime)
             continue
 
         # File line: three tokens if it describes a renaming, otherwise two
-        tokens = line.split('\t')
+        tokens = line.split("\t")
 
         # Possible statuses:
         # M: Modified (content changed)
@@ -411,7 +524,7 @@ def parse_log(filelist, dirlist, stats, git, merge=False, filterlist=None):
         # T: Type changed: to/from regular file, symlinks, submodules
         # R099: Renamed (moved), with % of unchanged content. 100 = pure rename
         # Not possible in log: C=Copied, U=Unmerged, X=Unknown, B=pairing Broken
-        status = tokens[0].split(' ')[-1]
+        status = tokens[0].split(" ")[-1]
         file = tokens[-1]
 
         # Handles non-ASCII chars and OS path separator
@@ -419,56 +532,76 @@ def parse_log(filelist, dirlist, stats, git, merge=False, filterlist=None):
 
         def do_file():
             if args.skip_older_than_commit and get_mtime_path(file) <= mtime:
-                stats['skip'] += 1
+                stats["skip"] += 1
                 return
             if args.debug:
-                log.debug("%d\t%d\t%d\t%s\t%s",
-                          stats['loglines'], stats['commits'], stats['files'],
-                          datestr, file)
+                log.debug(
+                    "%d\t%d\t%d\t%s\t%s",
+                    stats["loglines"],
+                    stats["commits"],
+                    stats["files"],
+                    datestr,
+                    file,
+                )
             try:
                 touch(os.path.join(git.workdir, file), mtime)
-                stats['touches'] += 1
+                stats["touches"] += 1
             except Exception as e:
                 log.error("ERROR: %s: %s", e, file)
-                stats['errors'] += 1
+                stats["errors"] += 1
 
         def do_dir():
             if args.debug:
-                log.debug("%d\t%d\t-\t%s\t%s",
-                          stats['loglines'], stats['commits'],
-                          datestr, "{}/".format(dirname or '.'))
+                log.debug(
+                    "%d\t%d\t-\t%s\t%s",
+                    stats["loglines"],
+                    stats["commits"],
+                    datestr,
+                    "{}/".format(dirname or "."),
+                )
             try:
                 touch(os.path.join(git.workdir, dirname), mtime)
-                stats['dirtouches'] += 1
+                stats["dirtouches"] += 1
             except Exception as e:
                 log.error("ERROR: %s: %s", e, dirname)
-                stats['direrrors'] += 1
+                stats["direrrors"] += 1
 
         if file in filelist:
-            stats['files'] -= 1
+            stats["files"] -= 1
             filelist.remove(file)
             do_file()
 
-        if args.dirs and status in ('A', 'D'):
+        if args.dirs and status in ("A", "D"):
             dirname = os.path.dirname(file)
             if dirname in dirlist:
                 dirlist.remove(dirname)
                 do_dir()
 
         # All files done?
-        if not stats['files']:
+        if not stats["files"]:
             git.terminate()
             return
 
 
 # Main Logic ##################################################################
 
+
 def main():
     start = time.time()  # yes, Wall time. CPU time is not realistic for users.
-    stats = {_: 0 for _ in ('loglines', 'commits', 'touches', 'skip', 'errors',
-                            'dirtouches', 'direrrors')}
+    stats = {
+        _: 0
+        for _ in (
+            "loglines",
+            "commits",
+            "touches",
+            "skip",
+            "errors",
+            "dirtouches",
+            "direrrors",
+        )
+    }
 
-    logging.basicConfig(level=args.loglevel, format='%(message)s')
+    logging.basicConfig(level=args.loglevel, format="%(message)s")
     log.trace("Arguments: %s", args)
 
     # First things first: Where and Who are we?
@@ -499,13 +632,16 @@ def main():
 
             # Symlink (to file, to dir or broken - git handles the same way)
             if not UPDATE_SYMLINKS and os.path.islink(fullpath):
-                log.warning("WARNING: Skipping symlink, no OS support for updates: %s",
-                            path)
+                log.warning(
+                    "WARNING: Skipping symlink, no OS support for updates: %s", path
+                )
                 continue
 
             # skip files which are older than given threshold
-            if (args.skip_older_than
-                    and start - get_mtime_path(fullpath) > args.skip_older_than):
+            if (
+                args.skip_older_than
+                and start - get_mtime_path(fullpath) > args.skip_older_than
+            ):
                 continue
 
             # Always add files relative to worktree root
@@ -519,15 +655,17 @@ def main():
     else:
         dirty = set(git.ls_dirty())
         if dirty:
-            log.warning("WARNING: Modified files in the working directory were ignored."
-                "\nTo include such files, commit your changes or use --force.")
+            log.warning(
+                "WARNING: Modified files in the working directory were ignored."
+                "\nTo include such files, commit your changes or use --force."
+            )
             filelist -= dirty
 
     # Build dir list to be processed
     dirlist = set(os.path.dirname(_) for _ in filelist) if args.dirs else set()
 
-    stats['totalfiles'] = stats['files'] = len(filelist)
-    log.info("{0:,} files to be processed in work dir".format(stats['totalfiles']))
+    stats["totalfiles"] = stats["files"] = len(filelist)
+    log.info("{0:,} files to be processed in work dir".format(stats["totalfiles"]))
 
     if not filelist:
         # Nothing to do. Exit silently and without errors, just like git does
@@ -544,10 +682,18 @@ def main():
         if args.missing and not args.merge:
             filterlist = list(filelist)
             missing = len(filterlist)
-            log.info("{0:,} files not found in log, trying merge commits".format(missing))
+            log.info(
+                "{0:,} files not found in log, trying merge commits".format(missing)
+            )
             for i in range(0, missing, STEPMISSING):
-                parse_log(filelist, dirlist, stats, git,
-                          merge=True, filterlist=filterlist[i:i + STEPMISSING])
+                parse_log(
+                    filelist,
+                    dirlist,
+                    stats,
+                    git,
+                    merge=True,
+                    filterlist=filterlist[i : i + STEPMISSING],
+                )
 
         # Still missing some?
         for file in filelist:
@@ -556,29 +702,33 @@ def main():
     # Final statistics
     # Suggestion: use git-log --before=mtime to brag about skipped log entries
     def log_info(msg, *a, width=13):
-        ifmt = '{:%d,}'    % (width,)  # not using 'n' for consistency with ffmt
-        ffmt = '{:%d,.2f}' % (width,)
+        ifmt = "{:%d,}" % (width,)  # not using 'n' for consistency with ffmt
+        ffmt = "{:%d,.2f}" % (width,)
         # %-formatting lacks a thousand separator, must pre-render with .format()
-        log.info(msg.replace('%d', ifmt).replace('%f', ffmt).format(*a))
+        log.info(msg.replace("%d", ifmt).replace("%f", ffmt).format(*a))
 
     log_info(
-        "Statistics:\n"
-        "%f seconds\n"
-        "%d log lines processed\n"
-        "%d commits evaluated",
-        time.time() - start, stats['loglines'], stats['commits'])
+        "Statistics:\n%f seconds\n%d log lines processed\n%d commits evaluated",
+        time.time() - start,
+        stats["loglines"],
+        stats["commits"],
+    )
 
     if args.dirs:
-        if stats['direrrors']: log_info("%d directory update errors", stats['direrrors'])
-        log_info("%d directories updated", stats['dirtouches'])
+        if stats["direrrors"]:
+            log_info("%d directory update errors", stats["direrrors"])
+        log_info("%d directories updated", stats["dirtouches"])
 
-    if stats['touches'] != stats['totalfiles']:
-                        log_info("%d files",              stats['totalfiles'])
-    if stats['skip']:   log_info("%d files skipped",      stats['skip'])
-    if stats['files']:  log_info("%d files missing",      stats['files'])
-    if stats['errors']: log_info("%d file update errors", stats['errors'])
+    if stats["touches"] != stats["totalfiles"]:
+        log_info("%d files", stats["totalfiles"])
+    if stats["skip"]:
+        log_info("%d files skipped", stats["skip"])
+    if stats["files"]:
+        log_info("%d files missing", stats["files"])
+    if stats["errors"]:
+        log_info("%d file update errors", stats["errors"])
 
-    log_info("%d files updated", stats['touches'])
+    log_info("%d files updated", stats["touches"])
 
     if args.test:
         log.info("TEST RUN - No files modified!")

.github/workflows/_release.yml

@@ -220,7 +220,7 @@ jobs:
         with:
           python-version: ${{ env.PYTHON_VERSION }}
 
-      - uses: actions/download-artifact@v4
+      - uses: actions/download-artifact@v5
         with:
           name: dist
           path: ${{ inputs.working-directory }}/dist/
@@ -379,7 +379,7 @@ jobs:
         with:
           python-version: ${{ env.PYTHON_VERSION }}
 
-      - uses: actions/download-artifact@v4
+      - uses: actions/download-artifact@v5
         if: startsWith(inputs.working-directory, 'libs/core')
         with:
           name: dist
@@ -447,7 +447,7 @@ jobs:
         with:
           python-version: ${{ env.PYTHON_VERSION }}
 
-      - uses: actions/download-artifact@v4
+      - uses: actions/download-artifact@v5
         with:
           name: dist
           path: ${{ inputs.working-directory }}/dist/
@@ -486,7 +486,7 @@ jobs:
         with:
           python-version: ${{ env.PYTHON_VERSION }}
 
-      - uses: actions/download-artifact@v4
+      - uses: actions/download-artifact@v5
         with:
           name: dist
           path: ${{ inputs.working-directory }}/dist/

.github/workflows/_test.yml

@@ -79,4 +79,4 @@ jobs:
           # grep will exit non-zero if the target message isn't found,
           # and `set -e` above will cause the step to fail.
           echo "$STATUS" | grep 'nothing to commit, working tree clean'
-          
+

.github/workflows/_test_pydantic.yml

@@ -64,4 +64,4 @@ jobs:
 
           # grep will exit non-zero if the target message isn't found,
           # and `set -e` above will cause the step to fail.
-          echo "$STATUS" | grep 'nothing to commit, working tree clean'
+          echo "$STATUS" | grep 'nothing to commit, working tree clean'

.github/workflows/_test_release.yml

@@ -85,7 +85,7 @@ jobs:
     steps:
       - uses: actions/checkout@v4
 
-      - uses: actions/download-artifact@v4
+      - uses: actions/download-artifact@v5
         with:
           name: test-dist
           path: ${{ inputs.working-directory }}/dist/

.github/workflows/api_doc_build.yml

@@ -52,7 +52,6 @@ jobs:
         run: |
           # Get unique repositories
           REPOS=$(echo "$REPOS_UNSORTED" | sort -u)
-
           # Checkout each unique repository
           for repo in $REPOS; do
             # Validate repository format (allow any org with proper format)
@@ -68,7 +67,6 @@ jobs:
               echo "Error: Invalid repository name: $REPO_NAME"
               exit 1
             fi
-
             echo "Checking out $repo to $REPO_NAME"
             git clone --depth 1 https://github.com/$repo.git $REPO_NAME
           done

.github/workflows/check_diffs.yml

@@ -30,6 +30,7 @@ jobs:
   build:
     name: 'Detect Changes & Set Matrix'
     runs-on: ubuntu-latest
+    if: ${{ !contains(github.event.pull_request.labels.*.name, 'ci-ignore') }}
     steps:
       - name: '📋 Checkout Code'
         uses: actions/checkout@v4

.github/workflows/codspeed.yml

@@ -20,6 +20,7 @@ jobs:
   codspeed:
     name: 'Benchmark'
     runs-on: ubuntu-latest
+    if: ${{ !contains(github.event.pull_request.labels.*.name, 'codspeed-ignore') }}
     strategy:
       matrix:
         include:

.markdownlint.json

@@ -11,4 +11,4 @@
   "MD046": {
     "style": "fenced"
   }
-}
+}

.vscode/settings.json

@@ -21,7 +21,7 @@
   "[python]": {
     "editor.formatOnSave": true,
     "editor.codeActionsOnSave": {
-      "source.organizeImports": "explicit",
+      "source.organizeImports.ruff": "explicit",
       "source.fixAll": "explicit"
     },
     "editor.defaultFormatter": "charliermarsh.ruff"
@@ -77,4 +77,6 @@
     "editor.tabSize": 2,
     "editor.insertSpaces": true
   },
+  "python.terminal.activateEnvironment": false,
+  "python.defaultInterpreterPath": "./.venv/bin/python"
 }

cookbook/README.md

@@ -63,4 +63,4 @@ Notebook | Description
 [rag-locally-on-intel-cpu.ipynb](https://github.com/langchain-ai/langchain/tree/master/cookbook/rag-locally-on-intel-cpu.ipynb) | Perform Retrieval-Augmented-Generation (RAG) on locally downloaded open-source models using langchain and open source tools and execute it on Intel Xeon CPU. We showed an example of how to apply RAG on Llama 2 model and enable it to answer the queries related to Intel Q1 2024 earnings release.
 [visual_RAG_vdms.ipynb](https://github.com/langchain-ai/langchain/tree/master/cookbook/visual_RAG_vdms.ipynb) | Performs Visual Retrieval-Augmented-Generation (RAG) using videos and scene descriptions generated by open source models.
 [contextual_rag.ipynb](https://github.com/langchain-ai/langchain/tree/master/cookbook/contextual_rag.ipynb) | Performs contextual retrieval-augmented generation (RAG) prepending chunk-specific explanatory context to each chunk before embedding.
-[rag-agents-locally-on-intel-cpu.ipynb](https://github.com/langchain-ai/langchain/tree/master/cookbook/local_rag_agents_intel_cpu.ipynb) | Build a RAG agent locally with open source models that routes questions through one of two paths to find answers. The agent generates answers based on documents retrieved from either the vector database or retrieved from web search. If the vector database lacks relevant information, the agent opts for web search. Open-source models for LLM and embeddings are used locally on an Intel Xeon CPU to execute this pipeline.
+[rag-agents-locally-on-intel-cpu.ipynb](https://github.com/langchain-ai/langchain/tree/master/cookbook/local_rag_agents_intel_cpu.ipynb) | Build a RAG agent locally with open source models that routes questions through one of two paths to find answers. The agent generates answers based on documents retrieved from either the vector database or retrieved from web search. If the vector database lacks relevant information, the agent opts for web search. Open-source models for LLM and embeddings are used locally on an Intel Xeon CPU to execute this pipeline.

docs/api_reference/create_api_rst.py

@@ -97,7 +97,7 @@ def _load_module_members(module_path: str, namespace: str) -> ModuleMembers:
             if type(type_) is typing_extensions._TypedDictMeta:  # type: ignore
                 kind: ClassKind = "TypedDict"
             elif type(type_) is typing._TypedDictMeta:  # type: ignore
-                kind: ClassKind = "TypedDict"
+                kind = "TypedDict"
             elif (
                 issubclass(type_, Runnable)
                 and issubclass(type_, BaseModel)
@@ -189,7 +189,7 @@ def _load_package_modules(
         if isinstance(package_directory, str)
         else package_directory
     )
-    modules_by_namespace = {}
+    modules_by_namespace: Dict[str, ModuleMembers] = {}
 
     # Get the high level package name
     package_name = package_path.name
@@ -217,7 +217,11 @@ def _load_package_modules(
         # Get the full namespace of the module
         namespace = str(relative_module_name).replace(".py", "").replace("/", ".")
         # Keep only the top level namespace
-        top_namespace = namespace.split(".")[0]
+        # (but make special exception for content_blocks and v1.messages)
+        if namespace == "messages.content_blocks" or namespace == "v1.messages":
+            top_namespace = namespace  # Keep full namespace for content_blocks
+        else:
+            top_namespace = namespace.split(".")[0]
 
         try:
             # If submodule is present, we need to construct the paths in a slightly
@@ -283,7 +287,7 @@ def _construct_doc(
 .. toctree::
     :hidden:
     :maxdepth: 2
-    
+
 """
     index_autosummary = """
 """
@@ -365,9 +369,9 @@ def _construct_doc(
 
                 module_doc += f"""\
     :template: {template}
-    
+
     {class_["qualified_name"]}
-    
+
 """
                 index_autosummary += f"""
     {class_["qualified_name"]}
@@ -550,8 +554,8 @@ def _build_index(dirs: List[str]) -> None:
     integrations = sorted(dir_ for dir_ in dirs if dir_ not in main_)
     doc = """# LangChain Python API Reference
 
-Welcome to the LangChain Python API reference. This is a reference for all 
-`langchain-x` packages. 
+Welcome to the LangChain Python API reference. This is a reference for all
+`langchain-x` packages.
 
 For user guides see [https://python.langchain.com](https://python.langchain.com).
 

docs/docs/concepts/async.mdx

@@ -1,4 +1,4 @@
-# Async programming with langchain
+# Async programming with LangChain
 
 :::info Prerequisites
 * [Runnable interface](/docs/concepts/runnables)
@@ -12,7 +12,7 @@ You are expected to be familiar with asynchronous programming in Python before r
 This guide specifically focuses on what you need to know to work with LangChain in an asynchronous context, assuming that you are already familiar with asynchronous programming.
 :::
 
-## Langchain asynchronous APIs
+## LangChain asynchronous APIs
 
 Many LangChain APIs are designed to be asynchronous, allowing you to build efficient and responsive applications.
 

docs/docs/concepts/tools.mdx

@@ -31,7 +31,7 @@ The key attributes that correspond to the tool's **schema**:
 The key methods to execute the function associated with the **tool**:
 
 - **invoke**: Invokes the tool with the given arguments.
-- **ainvoke**: Invokes the tool with the given arguments, asynchronously. Used for [async programming with Langchain](/docs/concepts/async).
+- **ainvoke**: Invokes the tool with the given arguments, asynchronously. Used for [async programming with LangChain](/docs/concepts/async).
 
 ## Create tools using the `@tool` decorator
 

docs/docs/contributing/how_to/testing.mdx

@@ -124,6 +124,47 @@ start "" htmlcov/index.html || open htmlcov/index.html
 
 ```
 
+## Snapshot Testing
+
+Some tests use [syrupy](https://github.com/tophat/syrupy) for snapshot testing, which captures the output of functions and compares them to stored snapshots. This is particularly useful for testing JSON schema generation and other structured outputs.
+
+### Updating Snapshots
+
+To update snapshots when the expected output has legitimately changed:
+
+```bash
+uv run --group test pytest path/to/test.py --snapshot-update
+```
+
+### Pydantic Version Compatibility Issues
+
+Pydantic generates different JSON schemas across versions, which can cause snapshot test failures in CI when tests run with different Pydantic versions than what was used to generate the snapshots.
+
+**Symptoms:**
+- CI fails with snapshot mismatches showing differences like missing or extra fields.
+- Tests pass locally but fail in CI with different Pydantic versions
+
+**Solution:**
+Locally update snapshots using the same Pydantic version that CI uses:
+
+1. **Identify the failing Pydantic version** from CI logs (e.g., `2.7.0`, `2.8.0`, `2.9.0`)
+
+2. **Update snapshots with that version:**
+   ```bash
+   uv run --with "pydantic==2.9.0" --group test pytest tests/unit_tests/path/to/test.py::test_name --snapshot-update
+   ```
+
+3. **Verify compatibility across supported versions:**
+   ```bash
+   # Test with the version you used to update
+   uv run --with "pydantic==2.9.0" --group test pytest tests/unit_tests/path/to/test.py::test_name
+
+   # Test with other supported versions
+   uv run --with "pydantic==2.8.0" --group test pytest tests/unit_tests/path/to/test.py::test_name
+   ```
+
+**Note:** Some tests use `@pytest.mark.skipif` decorators to only run with specific Pydantic version ranges (e.g., `PYDANTIC_VERSION_AT_LEAST_210`). Make sure to understand these constraints when updating snapshots.
+
 ## Coverage
 
 Code coverage (i.e. the amount of code that is covered by unit tests) helps identify areas of the code that are potentially more or less brittle.

docs/docs/how_to/graph_constructing.ipynb

@@ -122,13 +122,13 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 4,
+   "execution_count": null,
    "metadata": {},
    "outputs": [],
    "source": [
     "import os\n",
     "\n",
-    "from langchain_experimental.graph_transformers import LLMGraphTransformer\n",
+    "# from langchain_experimental.graph_transformers import LLMGraphTransformer\n",
     "from langchain_openai import ChatOpenAI\n",
     "\n",
     "llm = ChatOpenAI(temperature=0, model_name=\"gpt-4-turbo\")\n",

docs/docs/how_to/index.mdx

@@ -34,7 +34,7 @@ These are the core building blocks you can use when building applications.
 [Chat Models](/docs/concepts/chat_models) are newer forms of language models that take messages in and output a message.
 See [supported integrations](/docs/integrations/chat/) for details on getting started with chat models from a specific provider.
 
-- [How to: init any model in one line](/docs/how_to/chat_models_universal_init/)
+- [How to: initialize any model in one line](/docs/how_to/chat_models_universal_init/)
 - [How to: work with local models](/docs/how_to/local_llms)
 - [How to: do function/tool calling](/docs/how_to/tool_calling)
 - [How to: get models to return structured output](/docs/how_to/structured_output)

docs/docs/how_to/local_llms.ipynb

@@ -45,8 +45,8 @@
     "A few frameworks for this have emerged to support inference of open-source LLMs on various devices:\n",
     "\n",
     "1. [`llama.cpp`](https://github.com/ggerganov/llama.cpp): C++ implementation of llama inference code with [weight optimization / quantization](https://finbarr.ca/how-is-llama-cpp-possible/)\n",
-    "2. [`gpt4all`](https://docs.gpt4all.io/index.html): Optimized C backend for inference\n",
-    "3. [`Ollama`](https://ollama.ai/): Bundles model weights and environment into an app that runs on device and serves the LLM\n",
+    "2. [`gpt4all`](https://github.com/nomic-ai/gpt4all): Optimized C backend for inference\n",
+    "3. [`ollama`](https://github.com/ollama/ollama): Bundles model weights and environment into an app that runs on device and serves the LLM\n",
     "4. [`llamafile`](https://github.com/Mozilla-Ocho/llamafile): Bundles model weights and everything needed to run the model in a single file, allowing you to run the LLM locally from this file without any additional installation steps\n",
     "\n",
     "In general, these frameworks will do a few things:\n",
@@ -74,12 +74,12 @@
     "\n",
     "## Quickstart\n",
     "\n",
-    "[`Ollama`](https://ollama.ai/) is one way to easily run inference on macOS.\n",
+    "[Ollama](https://ollama.ai/) is one way to easily run inference on macOS.\n",
     " \n",
-    "The instructions [here](https://github.com/jmorganca/ollama?tab=readme-ov-file#ollama) provide details, which we summarize:\n",
+    "The instructions [here](https://github.com/ollama/ollama?tab=readme-ov-file#ollama) provide details, which we summarize:\n",
     " \n",
     "* [Download and run](https://ollama.ai/download) the app\n",
-    "* From command line, fetch a model from this [list of options](https://github.com/jmorganca/ollama): e.g., `ollama pull llama3.1:8b`\n",
+    "* From command line, fetch a model from this [list of options](https://ollama.com/search): e.g., `ollama pull llama3.1:8b`\n",
     "* When the app is running, all models are automatically served on `localhost:11434`\n"
    ]
   },
@@ -111,11 +111,11 @@
     }
    ],
    "source": [
-    "from langchain_ollama import OllamaLLM\n",
+    "from langchain_ollama import ChatOllama\n",
     "\n",
-    "llm = OllamaLLM(model=\"llama3.1:8b\")\n",
+    "llm = ChatOllama(model=\"gpt-oss:20b\")\n",
     "\n",
-    "llm.invoke(\"The first man on the moon was ...\")"
+    "llm.invoke(\"The first man on the moon was ...\").content"
    ]
   },
   {
@@ -149,40 +149,7 @@
    ],
    "source": [
     "for chunk in llm.stream(\"The first man on the moon was ...\"):\n",
-    "    print(chunk, end=\"|\", flush=True)"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "id": "e5731060",
-   "metadata": {},
-   "source": [
-    "Ollama also includes a chat model wrapper that handles formatting conversation turns:"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 4,
-   "id": "f14a778a",
-   "metadata": {},
-   "outputs": [
-    {
-     "data": {
-      "text/plain": [
-       "AIMessage(content='The answer is a historic one!\\n\\nThe first man to walk on the Moon was Neil Armstrong, an American astronaut and commander of the Apollo 11 mission. On July 20, 1969, Armstrong stepped out of the lunar module Eagle onto the surface of the Moon, famously declaring:\\n\\n\"That\\'s one small step for man, one giant leap for mankind.\"\\n\\nArmstrong was followed by fellow astronaut Edwin \"Buzz\" Aldrin, who also walked on the Moon during the mission. Michael Collins remained in orbit around the Moon in the command module Columbia.\\n\\nNeil Armstrong passed away on August 25, 2012, but his legacy as a pioneering astronaut and engineer continues to inspire people around the world!', response_metadata={'model': 'llama3.1:8b', 'created_at': '2024-08-01T00:38:29.176717Z', 'message': {'role': 'assistant', 'content': ''}, 'done_reason': 'stop', 'done': True, 'total_duration': 10681861417, 'load_duration': 34270292, 'prompt_eval_count': 19, 'prompt_eval_duration': 6209448000, 'eval_count': 141, 'eval_duration': 4432022000}, id='run-7bed57c5-7f54-4092-912c-ae49073dcd48-0', usage_metadata={'input_tokens': 19, 'output_tokens': 141, 'total_tokens': 160})"
-      ]
-     },
-     "execution_count": 4,
-     "metadata": {},
-     "output_type": "execute_result"
-    }
-   ],
-   "source": [
-    "from langchain_ollama import ChatOllama\n",
-    "\n",
-    "chat_model = ChatOllama(model=\"llama3.1:8b\")\n",
-    "\n",
-    "chat_model.invoke(\"Who was the first man on the moon?\")"
+    "    print(chunk.text(), end=\"|\", flush=True)"
    ]
   },
   {
@@ -200,7 +167,7 @@
     "\n",
     "### Running Apple silicon GPU\n",
     "\n",
-    "`Ollama` and [`llamafile`](https://github.com/Mozilla-Ocho/llamafile?tab=readme-ov-file#gpu-support) will automatically utilize the GPU on Apple devices.\n",
+    "`ollama` and [`llamafile`](https://github.com/Mozilla-Ocho/llamafile?tab=readme-ov-file#gpu-support) will automatically utilize the GPU on Apple devices.\n",
     " \n",
     "Other frameworks require the user to set up the environment to utilize the Apple GPU.\n",
     "\n",
@@ -212,15 +179,15 @@
     "\n",
     "In particular, ensure that conda is using the correct virtual environment that you created (`miniforge3`).\n",
     "\n",
-    "E.g., for me:\n",
+    "e.g.,\n",
     "\n",
-    "```\n",
+    "```shell\n",
     "conda activate /Users/rlm/miniforge3/envs/llama\n",
     "```\n",
     "\n",
     "With the above confirmed, then:\n",
     "\n",
-    "```\n",
+    "```shell\n",
     "CMAKE_ARGS=\"-DLLAMA_METAL=on\" FORCE_CMAKE=1 pip install -U llama-cpp-python --no-cache-dir\n",
     "```"
    ]
@@ -234,17 +201,13 @@
     "\n",
     "There are various ways to gain access to quantized model weights.\n",
     "\n",
-    "1. [`HuggingFace`](https://huggingface.co/TheBloke) - Many quantized model are available for download and can be run with framework such as [`llama.cpp`](https://github.com/ggerganov/llama.cpp). You can also download models in [`llamafile` format](https://huggingface.co/models?other=llamafile) from HuggingFace.\n",
-    "2. [`gpt4all`](https://gpt4all.io/index.html) - The model explorer offers a leaderboard of metrics and associated quantized models available for download \n",
-    "3. [`Ollama`](https://github.com/jmorganca/ollama) - Several models can be accessed directly via `pull`\n",
+    "1. [HuggingFace](https://huggingface.co/TheBloke) - Many quantized model are available for download and can be run with framework such as [`llama.cpp`](https://github.com/ggerganov/llama.cpp). You can also download models in [`llamafile` format](https://huggingface.co/models?other=llamafile) from HuggingFace.\n",
+    "2. [gpt4all](https://gpt4all.io/index.html) - The model explorer offers a leaderboard of metrics and associated quantized models available for download \n",
+    "3. [ollama](https://github.com/ollama/ollama) - Several models can be accessed directly via `pull`\n",
     "\n",
     "### Ollama\n",
     "\n",
-    "With [Ollama](https://github.com/jmorganca/ollama), fetch a model via `ollama pull <model family>:<tag>`:\n",
-    "\n",
-    "* E.g., for Llama 2 7b: `ollama pull llama2` will download the most basic version of the model (e.g., smallest # parameters and 4 bit quantization)\n",
-    "* We can also specify a particular version from the [model list](https://github.com/jmorganca/ollama?tab=readme-ov-file#model-library), e.g., `ollama pull llama2:13b`\n",
-    "* See the full set of parameters on the [API reference page](https://python.langchain.com/api_reference/community/llms/langchain_community.llms.ollama.Ollama.html)"
+    "With [Ollama](https://github.com/ollama/ollama), fetch a model via `ollama pull <model family>:<tag>`:"
    ]
   },
   {
@@ -265,7 +228,7 @@
     }
    ],
    "source": [
-    "llm = OllamaLLM(model=\"llama2:13b\")\n",
+    "llm = ChatOllama(model=\"gpt-oss:20b\")\n",
     "llm.invoke(\"The first man on the moon was ... think step by step\")"
    ]
   },
@@ -684,12 +647,6 @@
     "\n",
     "In addition, [here](https://blog.langchain.dev/using-langsmith-to-support-fine-tuning-of-open-source-llms/) is an overview on fine-tuning, which can utilize open-source LLMs."
    ]
-  },
-  {
-   "cell_type": "markdown",
-   "id": "14c2c170",
-   "metadata": {},
-   "source": []
   }
  ],
  "metadata": {

docs/docs/how_to/query_high_cardinality.ipynb

@@ -15,7 +15,7 @@
    "id": "f2195672-0cab-4967-ba8a-c6544635547d",
    "metadata": {},
    "source": [
-    "# How deal with high cardinality categoricals when doing query analysis\n",
+    "# How to deal with high-cardinality categoricals when doing query analysis\n",
     "\n",
     "You may want to do query analysis to create a filter on a categorical column. One of the difficulties here is that you usually need to specify the EXACT categorical value. The issue is you need to make sure the LLM generates that categorical value exactly. This can be done relatively easy with prompting when there are only a few values that are valid. When there are a high number of valid values then it becomes more difficult, as those values may not fit in the LLM context, or (if they do) there may be too many for the LLM to properly attend to.\n",
     "\n",

docs/docs/how_to/semantic-chunker.ipynb

@@ -74,12 +74,12 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 4,
+   "execution_count": null,
    "id": "a88ff70c",
    "metadata": {},
    "outputs": [],
    "source": [
-    "from langchain_experimental.text_splitter import SemanticChunker\n",
+    "# from langchain_experimental.text_splitter import SemanticChunker\n",
     "from langchain_openai.embeddings import OpenAIEmbeddings\n",
     "\n",
     "text_splitter = SemanticChunker(OpenAIEmbeddings())"

docs/docs/how_to/sql_csv.ipynb

@@ -612,56 +612,11 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 18,
+   "execution_count": null,
    "id": "35ea904e-795f-411b-bef8-6484dbb6e35c",
    "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "\n",
-      "\n",
-      "\u001b[1m> Entering new AgentExecutor chain...\u001b[0m\n",
-      "\u001b[32;1m\u001b[1;3m\n",
-      "Invoking: `python_repl_ast` with `{'query': \"df[['Age', 'Fare']].corr().iloc[0,1]\"}`\n",
-      "\n",
-      "\n",
-      "\u001b[0m\u001b[36;1m\u001b[1;3m0.11232863699941621\u001b[0m\u001b[32;1m\u001b[1;3m\n",
-      "Invoking: `python_repl_ast` with `{'query': \"df[['Fare', 'Survived']].corr().iloc[0,1]\"}`\n",
-      "\n",
-      "\n",
-      "\u001b[0m\u001b[36;1m\u001b[1;3m0.2561785496289603\u001b[0m\u001b[32;1m\u001b[1;3mThe correlation between Age and Fare is approximately 0.112, and the correlation between Fare and Survival is approximately 0.256.\n",
-      "\n",
-      "Therefore, the correlation between Fare and Survival (0.256) is greater than the correlation between Age and Fare (0.112).\u001b[0m\n",
-      "\n",
-      "\u001b[1m> Finished chain.\u001b[0m\n"
-     ]
-    },
-    {
-     "data": {
-      "text/plain": [
-       "{'input': \"What's the correlation between age and fare? is that greater than the correlation between fare and survival?\",\n",
-       " 'output': 'The correlation between Age and Fare is approximately 0.112, and the correlation between Fare and Survival is approximately 0.256.\\n\\nTherefore, the correlation between Fare and Survival (0.256) is greater than the correlation between Age and Fare (0.112).'}"
-      ]
-     },
-     "execution_count": 18,
-     "metadata": {},
-     "output_type": "execute_result"
-    }
-   ],
-   "source": [
-    "from langchain_experimental.agents import create_pandas_dataframe_agent\n",
-    "\n",
-    "agent = create_pandas_dataframe_agent(\n",
-    "    llm, df, agent_type=\"openai-tools\", verbose=True, allow_dangerous_code=True\n",
-    ")\n",
-    "agent.invoke(\n",
-    "    {\n",
-    "        \"input\": \"What's the correlation between age and fare? is that greater than the correlation between fare and survival?\"\n",
-    "    }\n",
-    ")"
-   ]
+   "outputs": [],
+   "source": "from langchain_experimental.agents import create_pandas_dataframe_agent\n\nagent = create_pandas_dataframe_agent(\n    llm, df, agent_type=\"openai-tools\", verbose=True, allow_dangerous_code=True\n)\nagent.invoke(\n    {\n        \"input\": \"What's the correlation between age and fare? is that greater than the correlation between fare and survival?\"\n    }\n)"
   },
   {
    "cell_type": "markdown",
@@ -786,4 +741,4 @@
  },
  "nbformat": 4,
  "nbformat_minor": 5
-}
+}

docs/docs/how_to/structured_output.ipynb

@@ -614,6 +614,7 @@
     "    HumanMessage(\"Now about caterpillars\", name=\"example_user\"),\n",
     "    AIMessage(\n",
     "        \"\",\n",
+    "        name=\"example_assistant\",\n",
     "        tool_calls=[\n",
     "            {\n",
     "                \"name\": \"joke\",\n",
@@ -909,7 +910,7 @@
     "        ),\n",
     "        (\"human\", \"{query}\"),\n",
     "    ]\n",
-    ").partial(schema=People.schema())\n",
+    ").partial(schema=People.model_json_schema())\n",
     "\n",
     "\n",
     "# Custom parser\n",

docs/docs/integrations/chat/ollama.ipynb

@@ -17,25 +17,29 @@
    "source": [
     "# ChatOllama\n",
     "\n",
-    "[Ollama](https://ollama.ai/) allows you to run open-source large language models, such as Llama 2, locally.\n",
+    "[Ollama](https://ollama.com/) allows you to run open-source large language models, such as `gpt-oss`, locally.\n",
     "\n",
-    "Ollama bundles model weights, configuration, and data into a single package, defined by a Modelfile.\n",
+    "`ollama` bundles model weights, configuration, and data into a single package, defined by a Modelfile.\n",
     "\n",
     "It optimizes setup and configuration details, including GPU usage.\n",
     "\n",
-    "For a complete list of supported models and model variants, see the [Ollama model library](https://github.com/jmorganca/ollama#model-library).\n",
+    "For a complete list of supported models and model variants, see the [Ollama model library](https://ollama.com/search).\n",
+    "\n",
+    ":::warning\n",
+    "This page is for the new v1 `ChatOllama` class with standard content block output. If you are looking for the legacy v0 `Ollama` class, see the [v0.3 documentation](https://python.langchain.com/v0.3/docs/integrations/chat/ollama/).\n",
+    ":::\n",
     "\n",
     "## Overview\n",
     "### Integration details\n",
     "\n",
-    "| Class | Package | Local | Serializable | [JS support](https://js.langchain.com/v0.2/docs/integrations/chat/ollama) | Package downloads | Package latest |\n",
+    "| Class | Package | Local | Serializable | [JS support](https://js.langchain.com/docs/integrations/chat/ollama/) | Package downloads | Package latest |\n",
     "| :--- | :--- | :---: | :---: |  :---: | :---: | :---: |\n",
-    "| [ChatOllama](https://python.langchain.com/v0.2/api_reference/ollama/chat_models/langchain_ollama.chat_models.ChatOllama.html) | [langchain-ollama](https://python.langchain.com/v0.2/api_reference/ollama/index.html) | ✅ | ❌ | ✅ | ![PyPI - Downloads](https://img.shields.io/pypi/dm/langchain-ollama?style=flat-square&label=%20) | ![PyPI - Version](https://img.shields.io/pypi/v/langchain-ollama?style=flat-square&label=%20) |\n",
+    "| [ChatOllama](https://python.langchain.com/api_reference/ollama/chat_models/langchain_ollama.chat_models.ChatOllama.html#chatollama) | [langchain-ollama](https://python.langchain.com/api_reference/ollama/index.html) | ✅ | ❌ | ✅ | ![PyPI - Downloads](https://img.shields.io/pypi/dm/langchain-ollama?style=flat-square&label=%20) | ![PyPI - Version](https://img.shields.io/pypi/v/langchain-ollama?style=flat-square&label=%20) |\n",
     "\n",
     "### Model features\n",
     "| [Tool calling](/docs/how_to/tool_calling/) | [Structured output](/docs/how_to/structured_output/) | JSON mode | [Image input](/docs/how_to/multimodal_inputs/) | Audio input | Video input | [Token-level streaming](/docs/how_to/chat_streaming/) | Native async | [Token usage](/docs/how_to/chat_token_usage_tracking/) | [Logprobs](/docs/how_to/logprobs/) |\n",
     "| :---: |:----------------------------------------------------:| :---: | :---: |  :---: | :---: | :---: | :---: | :---: | :---: |\n",
-    "| ✅ |                          ✅                           | ✅ | ❌ | ❌ | ❌ | ✅ | ✅ | ❌ | ❌ |\n",
+    "| ✅ |                          ✅                           | ✅ | ✅ | ❌ | ❌ | ✅ | ✅ | ❌ | ❌ |\n",
     "\n",
     "## Setup\n",
     "\n",
@@ -45,17 +49,17 @@
     "    * macOS users can install via Homebrew with `brew install ollama` and start with `brew services start ollama`\n",
     "* Fetch available LLM model via `ollama pull <name-of-model>`\n",
     "    * View a list of available models via the [model library](https://ollama.ai/library)\n",
-    "    * e.g., `ollama pull llama3`\n",
+    "    * e.g., `ollama pull gpt-oss:20b`\n",
     "* This will download the default tagged version of the model. Typically, the default points to the latest, smallest sized-parameter model.\n",
     "\n",
     "> On Mac, the models will be download to `~/.ollama/models`\n",
     ">\n",
     "> On Linux (or WSL), the models will be stored at `/usr/share/ollama/.ollama/models`\n",
     "\n",
-    "* Specify the exact version of the model of interest as such `ollama pull vicuna:13b-v1.5-16k-q4_0` (View the [various tags for the `Vicuna`](https://ollama.ai/library/vicuna/tags) model in this instance)\n",
+    "* Specify the exact version of the model of interest as such `ollama pull gpt-oss:20b`\n",
     "* To view all pulled models, use `ollama list`\n",
     "* To chat directly with a model from the command line, use `ollama run <name-of-model>`\n",
-    "* View the [Ollama documentation](https://github.com/ollama/ollama/tree/main/docs) for more commands. You can run `ollama help` in the terminal to see available commands.\n"
+    "* View the [Ollama documentation](https://github.com/ollama/ollama/blob/main/docs/README.md) for more commands. You can run `ollama help` in the terminal to see available commands.\n"
    ]
   },
   {
@@ -102,7 +106,11 @@
    "id": "b18bd692076f7cf7",
    "metadata": {},
    "source": [
-    "Make sure you're using the latest Ollama version for structured outputs. Update by running:"
+    ":::warning\n",
+    "Make sure you're using the latest Ollama client version!\n",
+    ":::\n",
+    "\n",
+    "Update by running:"
    ]
   },
   {
@@ -127,15 +135,16 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 9,
+   "execution_count": 2,
    "id": "cb09c344-1836-4e0c-acf8-11d13ac1dbae",
    "metadata": {},
    "outputs": [],
    "source": [
-    "from langchain_ollama import ChatOllama\n",
+    "from langchain_ollama.v1 import ChatOllama\n",
     "\n",
     "llm = ChatOllama(\n",
-    "    model=\"llama3.1\",\n",
+    "    model=\"gpt-oss:20b\",\n",
+    "    validate_model_on_init=True,\n",
     "    temperature=0,\n",
     "    # other params...\n",
     ")"
@@ -158,46 +167,56 @@
    },
    "outputs": [
     {
-     "data": {
-      "text/plain": [
-       "AIMessage(content='The translation of \"I love programming\" in French is:\\n\\n\"J\\'adore le programmation.\"', additional_kwargs={}, response_metadata={'model': 'llama3.1', 'created_at': '2025-06-25T18:43:00.483666Z', 'done': True, 'done_reason': 'stop', 'total_duration': 619971208, 'load_duration': 27793125, 'prompt_eval_count': 35, 'prompt_eval_duration': 36354583, 'eval_count': 22, 'eval_duration': 555182667, 'model_name': 'llama3.1'}, id='run--348bb5ef-9dd9-4271-bc7e-a9ddb54c28c1-0', usage_metadata={'input_tokens': 35, 'output_tokens': 22, 'total_tokens': 57})"
-      ]
-     },
-     "execution_count": 5,
-     "metadata": {},
-     "output_type": "execute_result"
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "AIMessage(type='ai', name=None, id='lc_run--5521db11-a5eb-4e46-956c-1455151cdaa3-0', lc_version='v1', content=[{'type': 'text', 'text': 'The translation of \"I love programming\" to French is:\\n\\n\"Je aime le programmation\"\\n\\nHowever, a more common and idiomatic way to express this in French would be:\\n\\n\"J\\'aime programmer\"\\n\\nThis phrase uses the verb \"aimer\" (to love) in the present tense, which is more suitable for expressing a general feeling or preference.'}], usage_metadata={'input_tokens': 34, 'output_tokens': 73, 'total_tokens': 107}, response_metadata={'model_name': 'llama3.2', 'created_at': '2025-08-08T23:07:44.439483Z', 'done': True, 'done_reason': 'stop', 'total_duration': 1410566833, 'load_duration': 28419542, 'prompt_eval_count': 34, 'prompt_eval_duration': 141642125, 'eval_count': 73, 'eval_duration': 1240075000}, parsed=None)\n",
+      "\n",
+      "Content:\n",
+      "The translation of \"I love programming\" to French is:\n",
+      "\n",
+      "\"Je aime le programmation\"\n",
+      "\n",
+      "However, a more common and idiomatic way to express this in French would be:\n",
+      "\n",
+      "\"J'aime programmer\"\n",
+      "\n",
+      "This phrase uses the verb \"aimer\" (to love) in the present tense, which is more suitable for expressing a general feeling or preference.\n"
+     ]
     }
    ],
    "source": [
-    "messages = [\n",
-    "    (\n",
-    "        \"system\",\n",
-    "        \"You are a helpful assistant that translates English to French. Translate the user sentence.\",\n",
-    "    ),\n",
-    "    (\"human\", \"I love programming.\"),\n",
-    "]\n",
-    "ai_msg = llm.invoke(messages)\n",
-    "ai_msg"
+    "ai_msg = llm.invoke(\"Translate 'I love programming' to French.\")\n",
+    "print(f\"{ai_msg}\\n\")\n",
+    "print(f\"Content:\\n{ai_msg.text}\")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "ede35e47",
+   "metadata": {},
+   "source": [
+    "## Streaming"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 11,
-   "id": "d86145b3-bfef-46e8-b227-4dda5c9c2705",
+   "execution_count": 10,
+   "id": "77474829",
    "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "The translation of \"I love programming\" in French is:\n",
-      "\n",
-      "\"J'adore le programmation.\"\n"
+      "Hi| there|!| I|'m| just| a| chat|bot|,| so| I| don|'t| have| feelings|,| but| I|'m| here| and| ready| to| help| you| with| anything| you| need|!| How| can| I| assist| you| today|?| 😊|"
      ]
     }
    ],
    "source": [
-    "print(ai_msg.content)"
+    "for chunk in llm.stream(\"How are you doing?\"):\n",
+    "    if chunk.text:\n",
+    "        print(chunk.text, end=\"|\", flush=True)"
    ]
   },
   {
@@ -219,10 +238,10 @@
     {
      "data": {
       "text/plain": [
-       "AIMessage(content='\"Programmieren ist meine Leidenschaft.\"\\n\\n(I translated \"programming\" to the German word \"Programmieren\", and added \"ist meine Leidenschaft\" which means \"is my passion\")', additional_kwargs={}, response_metadata={'model': 'llama3.1', 'created_at': '2025-06-25T18:43:29.350032Z', 'done': True, 'done_reason': 'stop', 'total_duration': 1194744459, 'load_duration': 26982500, 'prompt_eval_count': 30, 'prompt_eval_duration': 117043458, 'eval_count': 41, 'eval_duration': 1049892167, 'model_name': 'llama3.1'}, id='run--efc6436e-2346-43d9-8118-3c20b3cdf0d0-0', usage_metadata={'input_tokens': 30, 'output_tokens': 41, 'total_tokens': 71})"
+       "'Ich liebe Programmierung.'"
       ]
      },
-     "execution_count": 7,
+     "execution_count": 19,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -241,13 +260,15 @@
     ")\n",
     "\n",
     "chain = prompt | llm\n",
-    "chain.invoke(\n",
+    "result = chain.invoke(\n",
     "    {\n",
     "        \"input_language\": \"English\",\n",
     "        \"output_language\": \"German\",\n",
     "        \"input\": \"I love programming.\",\n",
     "    }\n",
-    ")"
+    ")\n",
+    "\n",
+    "result.text"
    ]
   },
   {
@@ -257,10 +278,10 @@
    "source": [
     "## Tool calling\n",
     "\n",
-    "We can use [tool calling](/docs/concepts/tool_calling/) with an LLM [that has been fine-tuned for tool use](https://ollama.com/search?&c=tools) such as `llama3.1`:\n",
+    "We can use [tool calling](/docs/concepts/tool_calling/) with an LLM [that has been fine-tuned for tool use](https://ollama.com/search?&c=tools) such as `gpt-oss`:\n",
     "\n",
     "```\n",
-    "ollama pull llama3.1\n",
+    "ollama pull gpt-oss:20b\n",
     "```\n",
     "\n",
     "Details on creating custom tools are available in [this guide](/docs/how_to/custom_tools/). Below, we demonstrate how to create a tool using the `@tool` decorator on a normal python function."
@@ -276,16 +297,16 @@
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "[{'name': 'validate_user', 'args': {'addresses': ['123 Fake St, Boston, MA', '234 Pretend Boulevard, Houston, TX'], 'user_id': '123'}, 'id': 'aef33a32-a34b-4b37-b054-e0d85584772f', 'type': 'tool_call'}]\n"
+      "[{'type': 'tool_call', 'id': 'f365489e-1dc4-4d60-aaff-e56290ae4f99', 'name': 'validate_user', 'args': {'addresses': ['123 Fake St in Boston MA', '234 Pretend Boulevard in Houston TX'], 'user_id': 123}}]\n"
      ]
     }
    ],
    "source": [
     "from typing import List\n",
     "\n",
-    "from langchain_core.messages import AIMessage\n",
+    "from langchain_core.v1.messages import AIMessage\n",
     "from langchain_core.tools import tool\n",
-    "from langchain_ollama import ChatOllama\n",
+    "from langchain_ollama.v1 import ChatOllama\n",
     "\n",
     "\n",
     "@tool\n",
@@ -300,7 +321,8 @@
     "\n",
     "\n",
     "llm = ChatOllama(\n",
-    "    model=\"llama3.1\",\n",
+    "    model=\"gpt-oss:20b\",\n",
+    "    validate_model_on_init=True,\n",
     "    temperature=0,\n",
     ").bind_tools([validate_user])\n",
     "\n",
@@ -314,6 +336,50 @@
     "    print(result.tool_calls)"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "id": "4321b6a8",
+   "metadata": {},
+   "source": [
+    "## Structured output"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 16,
+   "id": "20f8ae70",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Name: Alice, Age: 28, Job: Software Engineer\n"
+     ]
+    }
+   ],
+   "source": [
+    "from langchain_ollama.v1 import ChatOllama\n",
+    "from pydantic import BaseModel, Field\n",
+    "\n",
+    "llm = ChatOllama(model=\"llama3.2\", validate_model_on_init=True, temperature=0)\n",
+    "\n",
+    "\n",
+    "class Person(BaseModel):\n",
+    "    \"\"\"Information about a person.\"\"\"\n",
+    "\n",
+    "    name: str = Field(description=\"The person's full name\")\n",
+    "    age: int = Field(description=\"The person's age in years\")\n",
+    "    occupation: str = Field(description=\"The person's job or profession\")\n",
+    "\n",
+    "\n",
+    "structured_llm = llm.with_structured_output(Person)\n",
+    "response: Person = structured_llm.invoke(\n",
+    "    \"Tell me about a fictional software engineer named Alice who is 28 years old.\"\n",
+    ")\n",
+    "print(f\"Name: {response.name}, Age: {response.age}, Job: {response.occupation}\")"
+   ]
+  },
   {
    "cell_type": "markdown",
    "id": "4c5e0197",
@@ -321,11 +387,9 @@
    "source": [
     "## Multi-modal\n",
     "\n",
-    "Ollama has support for multi-modal LLMs, such as [bakllava](https://ollama.com/library/bakllava) and [llava](https://ollama.com/library/llava).\n",
+    "Ollama has limited support for multi-modal LLMs, such as [gemma3](https://ollama.com/library/gemma3).\n",
     "\n",
-    "    ollama pull bakllava\n",
-    "\n",
-    "Be sure to update Ollama so that you have the most recent version to support multi-modal."
+    "### Image input"
    ]
   },
   {
@@ -408,15 +472,15 @@
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "90%\n"
+      "Based on the image, the dollar-based gross retention rate is **90%**.\n"
      ]
     }
    ],
    "source": [
-    "from langchain_core.messages import HumanMessage\n",
-    "from langchain_ollama import ChatOllama\n",
+    "from langchain_core.v1.messages import HumanMessage\n",
+    "from langchain_ollama.v1 import ChatOllama\n",
     "\n",
-    "llm = ChatOllama(model=\"bakllava\", temperature=0)\n",
+    "llm = ChatOllama(model=\"gemma3:4b\", validate_model_on_init=True, temperature=0)\n",
     "\n",
     "\n",
     "def prompt_func(data):\n",
@@ -424,8 +488,9 @@
     "    image = data[\"image\"]\n",
     "\n",
     "    image_part = {\n",
-    "        \"type\": \"image_url\",\n",
-    "        \"image_url\": f\"data:image/jpeg;base64,{image}\",\n",
+    "        \"type\": \"image\",\n",
+    "        \"base64\": f\"data:image/jpeg;base64,{image}\",\n",
+    "        \"mime_type\": \"image/jpeg\",\n",
     "    }\n",
     "\n",
     "    content_parts = []\n",
@@ -435,7 +500,7 @@
     "    content_parts.append(image_part)\n",
     "    content_parts.append(text_part)\n",
     "\n",
-    "    return [HumanMessage(content=content_parts)]\n",
+    "    return [HumanMessage(content_parts)]\n",
     "\n",
     "\n",
     "from langchain_core.output_parsers import StrOutputParser\n",
@@ -454,11 +519,9 @@
    "id": "fb6a331f-1507-411f-89e5-c4d598154f3c",
    "metadata": {},
    "source": [
-    "## Reasoning models and custom message roles\n",
+    "## Reasoning models\n",
     "\n",
-    "Some models, such as IBM's [Granite 3.2](https://ollama.com/library/granite3.2), support custom message roles to enable thinking processes.\n",
-    "\n",
-    "To access Granite 3.2's thinking features, pass a message with a `\"control\"` role with content set to `\"thinking\"`. Because `\"control\"` is a non-standard message role, we can use a [ChatMessage](https://python.langchain.com/api_reference/core/messages/langchain_core.messages.chat.ChatMessage.html) object to implement it:"
+    "Many models support outputting their reasoning process in addition to the final answer. This is useful for debugging and understanding how the model arrived at its conclusion. This train of thought reasoning is available in models such as `gpt-oss`, `qwen3:8b`, and `deepseek-r1`. To enable reasoning output, set the `reasoning` parameter to `True` either when instantiating the model or during invocation."
    ]
   },
   {
@@ -471,30 +534,25 @@
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "Here is my thought process:\n",
-      "The user is asking for the value of 3 raised to the power of 3, which is a basic exponentiation operation.\n",
-      "\n",
-      "Here is my response:\n",
-      "\n",
-      "3^3 (read as \"3 to the power of 3\") equals 27. \n",
-      "\n",
-      "This calculation is performed by multiplying 3 by itself three times: 3*3*3 = 27.\n"
+      "Response including reasoning: [{'type': 'reasoning', 'reasoning': \"Okay, so I need to figure out what 3^3 is. Let me start by recalling what exponents mean. From what I remember, when you have a number raised to a power, like a^b, it means you multiply the number by itself b times. So, for example, 2^3 would be 2 multiplied by itself three times: 2 × 2 × 2. Let me check if that's right. Yeah, I think that's correct. So applying that to 3^3, it should be 3 multiplied by itself three times.\\n\\nWait, let me make sure I'm not confusing the base and the exponent. The base is the number being multiplied, and the exponent is how many times it's multiplied. So in 3^3, the base is 3 and the exponent is 3. That means I need to multiply 3 by itself three times. Let me write that out step by step.\\n\\nFirst, multiply the first two 3s: 3 × 3. What's 3 times 3? That's 9. Okay, so the first multiplication gives me 9. Now, I need to multiply that result by the third 3. So 9 × 3. Let me calculate that. 9 times 3 is... 27. So putting it all together, 3 × 3 × 3 equals 27. \\n\\nWait, let me verify that again. Maybe I should do it in a different way to make sure I didn't make a mistake. Let's break it down. 3^3 is the same as 3 × 3 × 3. Let me compute 3 × 3 first, which is 9, and then multiply that by 3. 9 × 3 is indeed 27. Hmm, that seems right. \\n\\nAlternatively, I can think of exponents as repeated multiplication. So 3^1 is 3, 3^2 is 3 × 3 = 9, and 3^3 is 3 × 3 × 3 = 27. Yeah, that progression makes sense. Each time the exponent increases by 1, you multiply by the base again. So starting from 3^1 = 3, then 3^2 is 3 × 3 = 9, then 3^3 is 9 × 3 = 27. \\n\\nIs there another way to check this? Maybe using exponent rules. For example, if I know that 3^2 is 9, then multiplying by another 3 would give me 3^3. Since 9 × 3 is 27, that confirms it again. \\n\\nAlternatively, maybe I can use logarithms or something else, but that might be overcomplicating. Since exponents are straightforward multiplication, I think my initial calculation is correct. \\n\\nWait, just to be thorough, maybe I can use a calculator to verify. Let me imagine pressing 3, then the exponent key, then 3. If I do that, it should give me 27. Yeah, that's what I remember. So all methods point to 27. \\n\\nI think I've checked it multiple ways: breaking down the multiplication step by step, using the exponent progression, and even considering a calculator verification. All of them lead to the same answer. Therefore, I'm confident that 3^3 equals 27.\\n\"}, {'type': 'text', 'text': 'To determine the value of $3^3$, we start by understanding what an exponent represents. The expression $a^b$ means multiplying the base $a$ by itself $b$ times. \\n\\n### Step-by-Step Calculation:\\n1. **Identify the base and exponent**:  \\n   In $3^3$, the base is **3**, and the exponent is **3**. This means we multiply 3 by itself three times.\\n\\n2. **Perform the multiplication**:  \\n   - First, multiply the first two 3s:  \\n     $3 \\\\times 3 = 9$  \\n   - Next, multiply the result by the third 3:  \\n     $9 \\\\times 3 = 27$\\n\\n3. **Verify the result**:  \\n   - $3^1 = 3$  \\n   - $3^2 = 3 \\\\times 3 = 9$  \\n   - $3^3 = 3 \\\\times 3 \\\\times 3 = 27$  \\n   This progression confirms the calculation.\\n\\n### Final Answer:\\n$$\\n3^3 = \\\\boxed{27}\\n$$'}]\n",
+      "Response without reasoning: [{'type': 'text', 'text': \"Sure! Let's break down what **3³** means and how to calculate it step by step.\\n\\n---\\n\\n### Step 1: Understand the notation\\nThe expression **3³** means **3 multiplied by itself three times**. The small number (3) is called the **exponent**, and it tells us how many times the base number (3) is used as a factor.\\n\\nSo:\\n$$\\n3^3 = 3 \\\\times 3 \\\\times 3\\n$$\\n\\n---\\n\\n### Step 2: Perform the multiplication step by step\\n\\n1. Multiply the first two 3s:\\n   $$\\n   3 \\\\times 3 = 9\\n   $$\\n\\n2. Now multiply the result by the third 3:\\n   $$\\n   9 \\\\times 3 = 27\\n   $$\\n\\n---\\n\\n### Step 3: Final Answer\\n\\n$$\\n3^3 = 27\\n$$\\n\\n---\\n\\n### Summary\\n- **3³** means **3 × 3 × 3**\\n- **3 × 3 = 9**\\n- **9 × 3 = 27**\\n- So, **3³ = 27**\\n\\nLet me know if you'd like to explore exponents further!\"}]\n"
      ]
     }
    ],
    "source": [
-    "from langchain_core.messages import ChatMessage, HumanMessage\n",
-    "from langchain_ollama import ChatOllama\n",
+    "from langchain_ollama.v1 import ChatOllama\n",
     "\n",
-    "llm = ChatOllama(model=\"granite3.2:8b\")\n",
+    "# All outputs from `llm` will include reasoning unless overridden during invocation\n",
+    "llm = ChatOllama(model=\"qwen3:8b\", validate_model_on_init=True, reasoning=True)\n",
     "\n",
-    "messages = [\n",
-    "    ChatMessage(role=\"control\", content=\"thinking\"),\n",
-    "    HumanMessage(\"What is 3^3?\"),\n",
-    "]\n",
+    "response_a = llm.invoke(\"What is 3^3? Explain your reasoning step by step.\")\n",
+    "print(f\"Response including reasoning: {response_a.content}\")\n",
     "\n",
-    "response = llm.invoke(messages)\n",
-    "print(response.content)"
+    "# Test override; note no ReasoningContentBlock in the response\n",
+    "response_b = llm.invoke(\n",
+    "    \"What is 3^3? Explain your reasoning step by step.\", reasoning=False\n",
+    ")\n",
+    "print(f\"Response without reasoning: {response_b.content}\")"
    ]
   },
   {
@@ -502,7 +560,7 @@
    "id": "6271d032-da40-44d4-9b52-58370e164be3",
    "metadata": {},
    "source": [
-    "Note that the model exposes its thought process in addition to its final response."
+    "Note that the model exposes its thought process as a `ReasoningContentBlock` addition to its final response."
    ]
   },
   {

docs/docs/integrations/chat/openai.ipynb

@@ -447,6 +447,163 @@
     ")"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "id": "c5d9d19d-8ab1-4d9d-b3a0-56ee4e89c528",
+   "metadata": {},
+   "source": [
+    "### Custom tools\n",
+    "\n",
+    ":::info Requires ``langchain-openai>=0.3.29``\n",
+    "\n",
+    ":::\n",
+    "\n",
+    "[Custom tools](https://platform.openai.com/docs/guides/function-calling#custom-tools) support tools with arbitrary string inputs. They can be particularly useful when you expect your string arguments to be long or complex."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "id": "a47c809b-852f-46bd-8b9e-d9534c17213d",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "================================\u001b[1m Human Message \u001b[0m=================================\n",
+      "\n",
+      "Use the tool to calculate 3^3.\n",
+      "==================================\u001b[1m Ai Message \u001b[0m==================================\n",
+      "\n",
+      "[{'id': 'rs_6894ff5747c0819d9b02fc5645b0be9c000169fd9fb68d99', 'summary': [], 'type': 'reasoning'}, {'call_id': 'call_7SYwMSQPbbEqFcKlKOpXeEux', 'input': 'print(3**3)', 'name': 'execute_code', 'type': 'custom_tool_call', 'id': 'ctc_6894ff5b9f54819d8155a63638d34103000169fd9fb68d99', 'status': 'completed'}]\n",
+      "Tool Calls:\n",
+      "  execute_code (call_7SYwMSQPbbEqFcKlKOpXeEux)\n",
+      " Call ID: call_7SYwMSQPbbEqFcKlKOpXeEux\n",
+      "  Args:\n",
+      "    __arg1: print(3**3)\n",
+      "=================================\u001b[1m Tool Message \u001b[0m=================================\n",
+      "Name: execute_code\n",
+      "\n",
+      "[{'type': 'custom_tool_call_output', 'output': '27'}]\n",
+      "==================================\u001b[1m Ai Message \u001b[0m==================================\n",
+      "\n",
+      "[{'type': 'text', 'text': '27', 'annotations': [], 'id': 'msg_6894ff5db3b8819d9159b3a370a25843000169fd9fb68d99'}]\n"
+     ]
+    }
+   ],
+   "source": [
+    "from langchain_openai import ChatOpenAI, custom_tool\n",
+    "from langgraph.prebuilt import create_react_agent\n",
+    "\n",
+    "\n",
+    "@custom_tool\n",
+    "def execute_code(code: str) -> str:\n",
+    "    \"\"\"Execute python code.\"\"\"\n",
+    "    return \"27\"\n",
+    "\n",
+    "\n",
+    "llm = ChatOpenAI(model=\"gpt-5\", output_version=\"responses/v1\")\n",
+    "\n",
+    "agent = create_react_agent(llm, [execute_code])\n",
+    "\n",
+    "input_message = {\"role\": \"user\", \"content\": \"Use the tool to calculate 3^3.\"}\n",
+    "for step in agent.stream(\n",
+    "    {\"messages\": [input_message]},\n",
+    "    stream_mode=\"values\",\n",
+    "):\n",
+    "    step[\"messages\"][-1].pretty_print()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "5ef93be6-6d4c-4eea-acfd-248774074082",
+   "metadata": {},
+   "source": [
+    "<details>\n",
+    "<summary>Context-free grammars</summary>\n",
+    "\n",
+    "OpenAI supports the specification of a [context-free grammar](https://platform.openai.com/docs/guides/function-calling#context-free-grammars) for custom tool inputs in `lark` or `regex` format. See [OpenAI docs](https://platform.openai.com/docs/guides/function-calling#context-free-grammars) for details. The `format` parameter can be passed into `@custom_tool` as shown below:"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "id": "2ae04586-be33-49c6-8947-7867801d868f",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "================================\u001b[1m Human Message \u001b[0m=================================\n",
+      "\n",
+      "Use the tool to calculate 3^3.\n",
+      "==================================\u001b[1m Ai Message \u001b[0m==================================\n",
+      "\n",
+      "[{'id': 'rs_689500828a8481a297ff0f98e328689c0681550c89797f43', 'summary': [], 'type': 'reasoning'}, {'call_id': 'call_jzH01RVhu6EFz7yUrOFXX55s', 'input': '3 * 3 * 3', 'name': 'do_math', 'type': 'custom_tool_call', 'id': 'ctc_6895008d57bc81a2b84d0993517a66b90681550c89797f43', 'status': 'completed'}]\n",
+      "Tool Calls:\n",
+      "  do_math (call_jzH01RVhu6EFz7yUrOFXX55s)\n",
+      " Call ID: call_jzH01RVhu6EFz7yUrOFXX55s\n",
+      "  Args:\n",
+      "    __arg1: 3 * 3 * 3\n",
+      "=================================\u001b[1m Tool Message \u001b[0m=================================\n",
+      "Name: do_math\n",
+      "\n",
+      "[{'type': 'custom_tool_call_output', 'output': '27'}]\n",
+      "==================================\u001b[1m Ai Message \u001b[0m==================================\n",
+      "\n",
+      "[{'type': 'text', 'text': '27', 'annotations': [], 'id': 'msg_6895009776b881a2a25f0be8507d08f20681550c89797f43'}]\n"
+     ]
+    }
+   ],
+   "source": [
+    "from langchain_openai import ChatOpenAI, custom_tool\n",
+    "from langgraph.prebuilt import create_react_agent\n",
+    "\n",
+    "grammar = \"\"\"\n",
+    "start: expr\n",
+    "expr: term (SP ADD SP term)* -> add\n",
+    "| term\n",
+    "term: factor (SP MUL SP factor)* -> mul\n",
+    "| factor\n",
+    "factor: INT\n",
+    "SP: \" \"\n",
+    "ADD: \"+\"\n",
+    "MUL: \"*\"\n",
+    "%import common.INT\n",
+    "\"\"\"\n",
+    "\n",
+    "format_ = {\"type\": \"grammar\", \"syntax\": \"lark\", \"definition\": grammar}\n",
+    "\n",
+    "\n",
+    "# highlight-next-line\n",
+    "@custom_tool(format=format_)\n",
+    "def do_math(input_string: str) -> str:\n",
+    "    \"\"\"Do a mathematical operation.\"\"\"\n",
+    "    return \"27\"\n",
+    "\n",
+    "\n",
+    "llm = ChatOpenAI(model=\"gpt-5\", output_version=\"responses/v1\")\n",
+    "\n",
+    "agent = create_react_agent(llm, [do_math])\n",
+    "\n",
+    "input_message = {\"role\": \"user\", \"content\": \"Use the tool to calculate 3^3.\"}\n",
+    "for step in agent.stream(\n",
+    "    {\"messages\": [input_message]},\n",
+    "    stream_mode=\"values\",\n",
+    "):\n",
+    "    step[\"messages\"][-1].pretty_print()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "c63430c9-c7b0-4e92-a491-3f165dddeb8f",
+   "metadata": {},
+   "source": [
+    "</details>"
+   ]
+  },
   {
    "cell_type": "markdown",
    "id": "84833dd0-17e9-4269-82ed-550639d65751",

docs/docs/integrations/graphs/kuzu_db.ipynb

@@ -132,12 +132,13 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 6,
+   "execution_count": null,
    "metadata": {},
    "outputs": [],
    "source": [
     "from langchain_core.documents import Document\n",
-    "from langchain_experimental.graph_transformers import LLMGraphTransformer\n",
+    "\n",
+    "# from langchain_experimental.graph_transformers import LLMGraphTransformer\n",
     "from langchain_openai import ChatOpenAI\n",
     "\n",
     "# Define the LLMGraphTransformer\n",

docs/docs/integrations/graphs/memgraph.ipynb

@@ -548,12 +548,12 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 14,
+   "execution_count": null,
    "metadata": {},
    "outputs": [],
    "source": [
     "from langchain_core.documents import Document\n",
-    "from langchain_experimental.graph_transformers import LLMGraphTransformer"
+    "# from langchain_experimental.graph_transformers import LLMGraphTransformer"
    ]
   },
   {

docs/docs/integrations/providers/gradientai.mdx

@@ -1,4 +1,4 @@
-# ChatGradient
+# DigitalOcean Gradient
 
 This will help you getting started with DigitalOcean Gradient [chat models](/docs/concepts/chat_models).
 

docs/docs/integrations/providers/ollama.mdx

@@ -1,14 +1,16 @@
 # Ollama
 
 >[Ollama](https://ollama.com/) allows you to run open-source large language models,
-> such as [Llama3.1](https://ai.meta.com/blog/meta-llama-3-1/), locally.
+> such as [gpt-oss](https://ollama.com/library/gpt-oss), locally.
 >
->`Ollama` bundles model weights, configuration, and data into a single package, defined by a Modelfile.
->It optimizes setup and configuration details, including GPU usage.
->For a complete list of supported models and model variants, see the [Ollama model library](https://ollama.ai/library).
+>The `ollama` [package](https://pypi.org/project/ollama/0.5.3/) bundles model weights,
+> configuration, and data into a single package, defined by a Modelfile. It optimizes
+> setup and configuration details, including GPU usage.
+>For a complete list of supported models and model variants, see the
+> [Ollama model library](https://ollama.com/search).
 
-See [this guide](/docs/how_to/local_llms) for more details
-on how to use `Ollama` with LangChain.
+See [this guide](/docs/how_to/local_llms/#ollama) for more details
+on how to use `ollama` with LangChain.
 
 ## Installation and Setup
 ### Ollama installation
@@ -23,10 +25,10 @@ Ollama will start as a background service automatically, if this is disabled, ru
 ollama serve
 ```
 
-After starting ollama, run `ollama pull <name-of-model>` to download a model from the [Ollama model library](https://ollama.ai/library):
+After starting ollama, run `ollama pull <name-of-model>` to download a model from the [Ollama model library](https://ollama.com/library):
 
 ```bash
-ollama pull llama3.1
+ollama pull gpt-oss:20b
 ```
 
 - This will download the default tagged version of the model. Typically, the default points to the latest, smallest sized-parameter model.

docs/docs/integrations/vectorstores/jaguar.ipynb

@@ -29,8 +29,8 @@
     "   Please refer to the instructions in:\n",
     "   [www.jaguardb.com](http://www.jaguardb.com)\n",
     "   For quick setup in docker environment:\n",
-    "   docker pull jaguardb/jaguardb_with_http\n",
-    "   docker run -d -p 8888:8888 -p 8080:8080 --name jaguardb_with_http  jaguardb/jaguardb_with_http\n",
+    "   docker pull jaguardb/jaguardb\n",
+    "   docker run -d -p 8888:8888 -p 8080:8080 --name jaguardb  jaguardb/jaguardb\n",
     "\n",
     "2. You must install the http client package for JaguarDB:\n",
     "   ```\n",

docs/docs/integrations/vectorstores/sqlserver.ipynb

@@ -591,7 +591,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 36,
+   "execution_count": null,
    "metadata": {
     "azdata_cell_guid": "d9127900-0942-48f1-bd4d-081c7fa3fcae",
     "language": "python"
@@ -606,7 +606,7 @@
     }
    ],
    "source": [
-    "from langchain.document_loaders import AzureBlobStorageFileLoader\n",
+    "from langchain_community.document_loaders import AzureBlobStorageFileLoader\n",
     "from langchain.text_splitter import RecursiveCharacterTextSplitter\n",
     "from langchain_core.documents import Document\n",
     "\n",

docs/docs/versions/v0_4/how_to_update.mdx

@@ -0,0 +1,107 @@
+---
+sidebar_position: 3
+---
+
+# How to update your code
+
+*Last updated: 08.08.25*
+
+If you maintain custom callbacks or output parsers, type checkers may raise errors if
+they do not accept the new message types as inputs. This guide describes how to
+address those issues.
+
+If you do not maintain custom callbacks or output parsers, there are no breaking
+changes. See our guide on the [new message types](/docs/versions/v0_4/messages) to learn
+about new features introduced in v0.4.
+
+## Custom callbacks
+
+[BaseCallbackHandler](https://python.langchain.com/api_reference/core/callbacks/langchain_core.callbacks.base.BaseCallbackHandler.html)
+now includes an attribute `accepts_new_messages` that defaults to False. When this
+attribute is False, the callback system in langchain-core will automatically convert
+new message types to old, so there should be no runtime errors. You can update callback
+signatures as below to fix type-checking errors:
+
+```python
+from langchain_core.v1.messages import AIMessage, AIMessageChunk, MessageV1
+
+def on_chat_model_start(
+    self,
+    serialized: dict[str, Any],
+    # highlight-next-line
+    messages: Union[list[list[BaseMessage]], list[MessageV1]],
+    *,
+    run_id: UUID,
+    parent_run_id: Optional[UUID] = None,
+    tags: Optional[list[str]] = None,
+    metadata: Optional[dict[str, Any]] = None,
+    **kwargs: Any,
+) -> Any:
+
+def on_llm_new_token(
+    self,
+    token: str,
+    *,
+    chunk: Optional[
+        # highlight-next-line
+        Union[GenerationChunk, ChatGenerationChunk, AIMessageChunk]
+    ] = None,
+    run_id: UUID,
+    parent_run_id: Optional[UUID] = None,
+    **kwargs: Any,
+) -> Any:
+
+def on_llm_end(
+    self,
+    # highlight-next-line
+    response: Union[LLMResult, AIMessage],
+    *,
+    run_id: UUID,
+    parent_run_id: Optional[UUID] = None,
+    **kwargs: Any,
+) -> Any:
+```
+You can also safely type-ignore mypy `override` errors here unless you switch
+`accepts_new_messages` to True.
+
+
+## Custom output parsers
+
+All output parsers in `langchain-core` have been updated to accept the new message
+types.
+
+If you maintain a custom output parser, `langchain-core` exposes a
+`convert_from_v1_message` function so that your parser can easily operate on the new
+message types:
+
+```python
+from langchain_core.messages.utils import convert_from_v1_message
+from langchain_core.v1.messages import AIMessage
+
+def parse_result(
+    self,
+    # highlight-next-line
+    result: Union[list[Generation], AIMessage],
+    *,
+    partial: bool = False,
+) -> Union[list[AgentAction], AgentFinish]:
+    # highlight-start
+    if isinstance(result, AIMessage):
+        result = [ChatGeneration(message=convert_from_v1_message(result))]
+    # highlight-end
+    ...
+
+def _transform(
+    # higlight-next-line
+    self, input: Iterator[Union[str, BaseMessage, AIMessage]]
+) -> Iterator[AddableDict]:
+    for chunk in input:
+        # higlight-start
+        if isinstance(chunk, AIMessage):
+            chunk = convert_from_v1_message(chunk)
+        # higlight-end
+        ...
+```
+This will allow your parser to work as before. You can also update the parser to
+natively handle the new message types to save this conversion step. See our guide on
+the [new message types](/docs/versions/v0_4/messages) for details.

docs/docs/versions/v0_4/index.mdx

@@ -0,0 +1,81 @@
+---
+sidebar_position: 1
+---
+
+# LangChain v0.4
+
+*Last updated: 08.08.25*
+
+## What's changed
+
+LangChain v0.4 allows developers to opt-in to new message types that will become default
+in LangChain v1.0. LangChain v1.0 will be released this fall. These messages provide
+fully typed, provider-agnostic content, introducing standard content blocks for
+reasoning, citations, server-side tool calls, and other LLM features. They also offer
+performance benefits over existing message classes.
+
+New message types have been added to a `v1` namespace in `langchain-core`. Select
+integration packages now also expose a `v1` namespace containing chat models that
+work with the new message types.
+
+Input types for callbacks and output parsers have been widened to accept the new message
+types. If you maintain custom callbacks or output parsers, type checkers may raise
+errors if they do not accept the new message types as inputs. Refer to
+[this guide](/docs/versions/v0_4/how_to_update) for how to address those issues. This
+is the only breaking change.
+
+## What's new
+
+You can access the new chat models through [init_chat_model](/docs/how_to/chat_models_universal_init/) by setting `message_version="v1"`:
+
+```python
+from langchain.chat_models import init_chat_model
+
+llm = init_chat_model("openai:gpt-5", message_version="v1")
+
+input_message = {"role": "user", "content": "Hello, world!"}
+llm.invoke([input_message])
+```
+
+You can also access the `v1` namespaces directly:
+```python
+from langchain_core.v1.messages import HumanMessage
+from langchain_openai.v1 import ChatOpenAI
+
+input_message = HumanMessage("Hello, world!")
+llm.invoke([input_message])
+```
+
+:::info New message details
+
+See our guide on the [new message types](/docs/versions/v0_4/messages) for details.
+
+:::
+
+## How to update your code
+
+If you maintain custom callbacks or output parsers, type checkers may raise errors if
+they do not accept the new message types as inputs. Refer to
+[this guide](/docs/versions/v0_4/how_to_update) for how to address those issues.
+
+If you do not maintain custom callbacks or output parsers, there are no breaking
+changes. See our guide on the [new message types](/docs/versions/v0_4/messages) to learn
+about new features introduced in v0.4.
+
+
+
+### Base packages
+
+| Package                  | Latest | Recommended constraint |
+|--------------------------|--------|------------------------|
+| langchain                | 0.4.0  | >=0.4,<1.0             |
+| langchain-community      | 0.4.0  | >=0.4,<1.0             |
+| langchain-text-splitters | 0.4.0  | >=0.4,<1.0             |
+| langchain-core           | 0.4.0  | >=0.4,<1.0             |
+| langchain-experimental   | 0.4.0  | >=0.4,<1.0             |
+
+
+### Integration packages
+
+...
+

docs/docs/versions/v0_4/messages.mdx

@@ -0,0 +1,429 @@
+---
+sidebar_position: 2
+---
+
+# LangChain v1.0 message types
+
+*Last updated: 08.08.25*
+
+LangChain v0.4 allows developers to opt-in to new message types that will become default
+in LangChain v1.0. LangChain v1.0 will be released this fall.
+
+These messages should be considered a beta feature and are subject to change in
+LangChain v1.0, although we do not anticipate any significant changes.
+
+## Benefits
+
+The new message types offer improvements in performance, type-safety, and consistency
+across OpenAI, Anthropic, Gemini, and other providers.
+
+### Performance
+
+Importantly, the new messages are Python dataclasses, saving some runtime from
+instantiating (layers of) Pydantic BaseModels.
+
+LangChain v0.4 introduces a new `BaseChatModel` class in `langchain_core.v1.chat_models`
+that is faster and leaner than the existing `BaseChatModel` class, offering significant
+reductions in overhead above provider SDKs.
+
+### Type-safety
+
+Message content is typed as
+```python
+import langchain_core.messages.content_blocks as types
+
+content: list[types.ContentBlock]
+```
+
+where we have introduced standard types for text, reasoning, citations, server-side
+tool executions (e.g., web search and code interpreters). These include
+[tool calls](https://python.langchain.com/docs/concepts/tool_calling/) and the
+[multi-modal types](/docs/how_to/multimodal_inputs/) introduced in earlier versions
+of LangChain. There are no breaking changes associated with the existing content types.
+
+**This is the most significant change from the existing message classes**, which permit
+strings, lists of strings, or lists of untyped dicts as content. We have added a
+`.text` getter so that developers can easily recover string content. Consequently, we
+have deprecated `.text()` (as a method) in favor of the new property.
+
+`.tool_calls`, instead of an attribute, is now also a getter with an associated setter,
+so that usage is largely the same. See [usage comparison](#usage-comparison), below,
+for details.
+
+### Consistency
+
+Many chat models can generate a variety of content in a single conversational turn,
+including reasoning, tool calls and responses, images, text with citations, and other
+structured objects. We have standardized these types, resulting in improved
+inter-operability of messages across models.
+
+## Usage comparison
+
+| Task                    | Previous                               | New                                                              |
+|-------------------------|----------------------------------------|------------------------------------------------------------------|
+| Get text content (str)  | `message.content` or `message.text()`  | `message.text`                                                   |
+| Get content blocks      | `message.content`                      | `message.content`                                                |
+| Get `additional_kwargs` | `message.additional_kwargs`            | `[block for block in message.content if block["type"] == "..."]` |
+
+Getting `response_metadata` and `tool_calls` has not changed.
+
+### Changes in content blocks
+
+For providers that generate `list[dict]` content, the dict elements have changed to
+conform to the new content block types. Refer to the
+[API reference](https://python.langchain.com/api_reference/core/messages.html) for
+details. Below we show some examples.
+
+Importantly:
+- Where provider-specific fields map to fields on standard types, LangChain manages
+the translation.
+- Where provider-specific fields do not map to fields on standard types, LangChain
+stores them in an `"extras"` key (see below for examples).
+
+<details>
+  <summary>Citations and web search</summary>
+
+<div className="row">
+  <div className="col col--6" style={{minWidth: 0}}>
+    **Old content**
+```python
+from langchain.chat_models import init_chat_model
+
+llm = init_chat_model("openai:gpt-5-mini", output_version="responses/v1")
+llm_with_tools = llm.bind_tools([{"type": "web_search_preview"}])
+
+response = llm_with_tools.invoke("What was a positive news story from today?")
+response.content
+```
+```
+[
+  {
+    "type": "reasoning",
+    "id": "rs_abc123",
+    "summary": []
+  },
+  {
+    "type": "web_search_call",
+    "id": "ws_abc123",
+    "action": {
+      "query": "positive news today August 8 2025 'good news' 'Aug 8 2025' 'today' ",
+      "type": "search"
+    },
+    "status": "completed"
+  },
+  {
+    "type": "text",
+    "text": "Here are two positive news items from today...",
+    "annotations": [
+      {
+        "type": "url_citation",
+        "end_index": 455,
+        "start_index": 196,
+        "title": "Document title",
+        "url": "<document url>"
+      },
+      {
+        "type": "url_citation",
+        "end_index": 1022,
+        "start_index": 707,
+        "title": "Another Document",
+        "url": "<another document url>"
+      },
+    ],
+    "id": "msg_abc123"
+  }
+]
+```
+</div>
+
+  <div className="col col--6" style={{minWidth: 0}}>
+    **New content**
+```python
+from langchain.chat_models import init_chat_model
+
+llm = init_chat_model("openai:gpt-5-mini", message_version="v1")
+llm_with_tools = llm.bind_tools([{"type": "web_search_preview"}])
+
+response = llm_with_tools.invoke("What was a positive news story from today?")
+response.content
+```
+```
+[
+  {
+    "type": "reasoning",
+    "id": "rs_abc123"
+  },
+  {
+    "type": "web_search_call",
+    "id": "ws_abc123",
+    "query": "positive news August 8 2025 'good news' 'today' ",
+    "extras": {
+      "action": {"type": "search"},
+      "status": "completed",
+    }
+  },
+  {
+    "type": "web_search_result",
+    "id": "ws_abc123"
+  },
+  {
+    "type": "text",
+    "text": "Here are two positive news items from today...",
+    "annotations": [
+      {
+        "type": "citation",
+        "end_index": 455,
+        "start_index": 196,
+        "title": "Document title",
+        "url": "<document url>"
+      },
+      {
+        "type": "citation",
+        "end_index": 1022,
+        "start_index": 707,
+        "title": "Another Document",
+        "url": "<another document url>"
+      }
+    ],
+    "id": "msg_abc123"
+  }
+]
+```
+  </div>
+</div>
+</details>
+
+
+<details>
+  <summary>Reasoning</summary>
+
+<div className="row">
+  <div className="col col--6" style={{minWidth: 0}}>
+    **Old content**
+```python
+from langchain.chat_models import init_chat_model
+
+llm = init_chat_model(
+    "openai:gpt-5",
+    reasoning={"effort": "medium", "summary": "auto"},
+    output_version="responses/v1",
+)
+response = llm.invoke(
+    "What was the third tallest building in the world in the year 2000?"
+)
+response.content
+```
+```
+[
+  {
+    "type": "reasoning",
+    "id": "rs_abc123",
+    "summary": [
+      {
+        "text": "The user is asking about...",
+        "type": "summary_text"
+      },
+      {
+        "text": "We should consider...",
+        "type": "summary_text"
+      }
+    ]
+  },
+  {
+    "type": "text",
+    "text": "In the year 2000 the third-tallest building in the world was...",
+    "id": "msg_abc123"
+  }
+]
+```
+  </div>
+
+  <div className="col col--6" style={{minWidth: 0}}>
+    **New content**
+```python
+from langchain.chat_models import init_chat_model
+
+llm = init_chat_model(
+    "openai:gpt-5",
+    reasoning={"effort": "medium", "summary": "auto"},
+    message_version="v1",
+)
+response = llm.invoke(
+    "What was the third tallest building in the world in the year 2000?"
+)
+response.content
+```
+```
+[
+  {
+    "type": "reasoning",
+    "reasoning": "The user is asking about...",
+    "id": "rs_abc123"
+  },
+  {
+    "type": "reasoning",
+    "reasoning": "We should consider...",
+    "id": "rs_abc123"
+  },
+  {
+    "type": "text",
+    "text": "In the year 2000 the third-tallest building in the world was...",
+    "id": "msg_abc123"
+  }
+]
+```
+  </div>
+</div>
+</details>
+
+
+<details>
+  <summary>Non-standard blocks</summary>
+
+Where content blocks from specific providers do not map to a standard type, they are
+structured into a `"non_standard"` block:
+```python
+{
+    "type": "non_standard",
+    "value": original_block,
+}
+```
+<div className="row">
+  <div className="col col--6" style={{minWidth: 0}}>
+    **Old content**
+```python
+from langchain.chat_models import init_chat_model
+
+llm = init_chat_model("openai:gpt-5-mini", output_version="responses/v1")
+llm_with_tools = llm.bind_tools(
+    [
+        {
+            "type": "file_search",
+            "vector_store_ids": ["vs_67d0baa0544c8191be194a85e19cbf92"],
+        }
+    ]
+)
+
+response = llm_with_tools.invoke("What is deep research by OpenAI?")
+response.content
+```
+```
+[
+  {
+    "type": "reasoning",
+    "id": "rs_abc123",
+    "summary": []
+  },
+  {
+    "type": "file_search_call",
+    "id": "fs_abc123",
+    "queries": [
+      "What is deep research by OpenAI?",
+      "deep research OpenAI definition"
+    ],
+    "status": "completed"
+  },
+  {
+    "type": "reasoning",
+    "id": "rs_def456",
+    "summary": []
+  },
+  {
+    "type": "text",
+    "text": "Deep research is...",
+    "annotations": [
+      {
+        "type": "file_citation",
+        "file_id": "file-abc123",
+        "filename": "sample_file.pdf",
+        "index": 305
+      },
+      {
+        "type": "file_citation",
+        "file_id": "file-abc123",
+        "filename": "sample_file.pdf",
+        "index": 675
+      },
+    ],
+    "id": "msg_abc123"
+  }
+]
+```
+</div>
+
+  <div className="col col--6" style={{minWidth: 0}}>
+    **New content**
+```python
+from langchain.chat_models import init_chat_model
+
+llm = init_chat_model("openai:gpt-5-mini", message_version="v1")
+llm_with_tools = llm.bind_tools(
+    [
+        {
+            "type": "file_search",
+            "vector_store_ids": ["vs_67d0baa0544c8191be194a85e19cbf92"],
+        }
+    ]
+)
+
+response = llm_with_tools.invoke("What is deep research by OpenAI?")
+response.content
+```
+```
+[
+  {
+    "type": "reasoning",
+    "id": "rs_abc123",
+    "summary": []
+  },
+  {
+    "type": "non_standard",
+    "value": {
+      "type": "file_search_call",
+      "id": "fs_abc123",
+      "queries": [
+        "What is deep research by OpenAI?",
+        "deep research OpenAI definition"
+      ],
+      "status": "completed"
+    }
+  },
+  {
+    "type": "reasoning",
+    "id": "rs_def456",
+    "summary": []
+  },
+  {
+    "type": "text",
+    "text": "Deep research is...",
+    "annotations": [
+      {
+        "type": "citation",
+        "title": "sample_file.pdf",
+        "extras": {
+          "file_id": "file-abc123",
+          "index": 305
+        }
+      },
+      {
+        "type": "citation",
+        "title": "sample_file.pdf",
+        "extras": {
+          "file_id": "file-abc123",
+          "index": 675
+        }
+      },
+    ],
+    "id": "msg_abc123"
+  }
+]
+```
+  </div>
+</div>
+</details>
+
+
+## Feature gaps
+
+The new message types do not yet support LangChain's caching layer. Support will be
+added in the coming weeks.

docs/docusaurus.config.js

@@ -224,13 +224,17 @@ const config = {
           },
           {
             type: "dropdown",
-            label: "v0.3",
+            label: "v0.4",
             position: "right",
             items: [
               {
-                label: "v0.3",
+                label: "v0.4",
                 href: "/docs/introduction",
               },
+              {
+                label: "v0.3",
+                href: "https://python.langchain.com/v0.3/docs/introduction/",
+              },
               {
                 label: "v0.2",
                 href: "https://python.langchain.com/v0.2/docs/introduction",

docs/scripts/packages_yml_get_downloads.py

@@ -1,5 +1,6 @@
 from datetime import datetime, timedelta, timezone
 from pathlib import Path
+import re
 
 import requests
 from ruamel.yaml import YAML
@@ -11,10 +12,18 @@ PACKAGE_YML = Path(__file__).parents[2] / "libs" / "packages.yml"
 
 
 def _get_downloads(p: dict) -> int:
-    url = f"https://pypistats.org/api/packages/{p['name']}/recent?period=month"
-    r = requests.get(url)
-    r.raise_for_status()
-    return r.json()["data"]["last_month"]
+    url = f"https://pepy.tech/badge/{p['name']}/month"
+    svg = requests.get(url, timeout=10).text
+    texts = re.findall(r"<text[^>]*>([^<]+)</text>", svg)
+    latest = texts[-1].strip() if texts else "0"
+
+    # parse "1.2k", "3.4M", "12,345" -> int
+    latest = latest.replace(",", "")
+    if latest.endswith(("k", "K")):
+        return int(float(latest[:-1]) * 1_000)
+    if latest.endswith(("m", "M")):
+        return int(float(latest[:-1]) * 1_000_000)
+    return int(float(latest) if "." in latest else int(latest))
 
 
 current_datetime = datetime.now(timezone.utc)

docs/scripts/partner_pkg_table.py

@@ -101,7 +101,12 @@ def package_row(p: dict) -> str:
     link = p["provider_page"]
     title = p["name_title"]
     provider = f"[{title}]({link})" if link else title
-    return f"| {provider} | [{p['name']}]({p['package_url']}) | ![PyPI - Downloads](https://img.shields.io/pypi/dm/{p['name']}?style=flat-square&label=%20&color=blue) | ![PyPI - Version](https://img.shields.io/pypi/v/{p['name']}?style=flat-square&label=%20&color=orange) | {js} |"
+    return (
+        f"| {provider} | [{p['name']}]({p['package_url']}) | "
+        f"![Downloads](https://static.pepy.tech/badge/{p['name']}/month) | "
+        f"![PyPI - Version](https://img.shields.io/pypi/v/{p['name']}?style=flat-square&label=%20&color=orange) | "
+        f"{js} |"
+    )
 
 
 def table() -> str:

docs/sidebars.js

@@ -82,6 +82,14 @@ module.exports = {
       collapsed: false,
       collapsible: false,
       items: [
+        {
+          type: "category",
+          label: "v0.4",
+          items: [{
+            type: 'autogenerated',
+            dirName: 'versions/v0_4',
+          }],
+        },
         {
           type: 'doc',
           id: 'versions/v0_3/index',
@@ -418,7 +426,7 @@ module.exports = {
             },
           ],
         },
-        
+
       ],
       link: {
         type: "generated-index",

docs/src/theme/ChatModelTabs.js

@@ -231,6 +231,13 @@ ${llmVarName} = ChatWatsonx(
       model: "llama-3.1-sonar-small-128k-online",
       apiKeyName: "PPLX_API_KEY",
       packageName: "langchain-perplexity",
+    },
+    {
+      value: "deepseek",
+      label: "DeepSeek",
+      model: "deepseek-chat",
+      apiKeyName: "DEEPSEEK_API_KEY",
+      packageName: "langchain-deepseek",
     }
   ].map((item) => ({
     ...item,

docs/vercel.json

@@ -23,11 +23,19 @@
     {
       "source": "/v0.2/:path(.*/?)*",
       "destination": "https://langchain-v02.vercel.app/v0.2/:path*"
+    },
+    {
+      "source": "/v0.3",
+      "destination": "https://langchain-v03.vercel.app/v0.3"
+    },
+    {
+      "source": "/v0.3/:path(.*/?)*",
+      "destination": "https://langchain-v03.vercel.app/v0.3/:path*"
     }
   ],
   "redirects": [
     {
-      "source": "/v0.3/docs/:path(.*/?)*",
+      "source": "/v0.4/docs/:path(.*/?)*",
       "destination": "/docs/:path*"
     },
     {

libs/cli/langchain_cli/integration_template/README.md

@@ -35,6 +35,7 @@ embeddings.embed_query("What is the meaning of life?")
 ```
 
 ## LLMs
+
 `__ModuleName__LLM` class exposes LLMs from __ModuleName__.
 
 ```python

libs/cli/langchain_cli/namespaces/migrate/.grit/grit.yaml

@@ -1,3 +1,3 @@
 version: 0.0.1
 patterns:
-  - name: github.com/getgrit/stdlib#*
+  - name: github.com/getgrit/stdlib#*

libs/cli/langchain_cli/package_template/README.md

@@ -27,16 +27,16 @@ langchain app add __package_name__
 ```
 
 And add the following code to your `server.py` file:
+
 ```python
 __app_route_code__
 ```
 
-(Optional) Let's now configure LangSmith. 
-LangSmith will help us trace, monitor and debug LangChain applications. 
-You can sign up for LangSmith [here](https://smith.langchain.com/). 
+(Optional) Let's now configure LangSmith.
+LangSmith will help us trace, monitor and debug LangChain applications.
+You can sign up for LangSmith [here](https://smith.langchain.com/).
 If you don't have access, you can skip this section
 
-
 ```shell
 export LANGSMITH_TRACING=true
 export LANGSMITH_API_KEY=<your-api-key>
@@ -49,11 +49,11 @@ If you are inside this directory, then you can spin up a LangServe instance dire
 langchain serve
 ```
 
-This will start the FastAPI app with a server is running locally at 
+This will start the FastAPI app with a server is running locally at
 [http://localhost:8000](http://localhost:8000)
 
 We can see all templates at [http://127.0.0.1:8000/docs](http://127.0.0.1:8000/docs)
-We can access the playground at [http://127.0.0.1:8000/__package_name__/playground](http://127.0.0.1:8000/__package_name__/playground)  
+We can access the playground at [http://127.0.0.1:8000/__package_name__/playground](http://127.0.0.1:8000/__package_name__/playground)
 
 We can access the template from code with:
 
@@ -61,4 +61,4 @@ We can access the template from code with:
 from langserve.client import RemoteRunnable
 
 runnable = RemoteRunnable("http://localhost:8000/__package_name__")
-```
+```

libs/cli/langchain_cli/project_template/README.md

@@ -11,7 +11,7 @@ pip install -U langchain-cli
 ## Adding packages
 
 ```bash
-# adding packages from 
+# adding packages from
 # https://github.com/langchain-ai/langchain/tree/master/templates
 langchain app add $PROJECT_NAME
 
@@ -31,10 +31,10 @@ langchain app remove my/custom/path/rag
 ```
 
 ## Setup LangSmith (Optional)
-LangSmith will help us trace, monitor and debug LangChain applications. 
-You can sign up for LangSmith [here](https://smith.langchain.com/). 
-If you don't have access, you can skip this section
 
+LangSmith will help us trace, monitor and debug LangChain applications.
+You can sign up for LangSmith [here](https://smith.langchain.com/).
+If you don't have access, you can skip this section
 
 ```shell
 export LANGSMITH_TRACING=true

libs/core/langchain_core/_api/beta_decorator.py

@@ -144,10 +144,9 @@ def beta(
                 obj.__init__ = functools.wraps(obj.__init__)(  # type: ignore[misc]
                     warn_if_direct_instance
                 )
-                return cast("T", obj)
+                return obj
 
         elif isinstance(obj, property):
-            # note(erick): this block doesn't seem to be used?
             if not _obj_type:
                 _obj_type = "attribute"
             wrapped = None
@@ -168,6 +167,7 @@ def beta(
                     self.__orig_fget = fget
                     self.__orig_fset = fset
                     self.__orig_fdel = fdel
+                    self.__doc__ = doc
 
                 def __get__(
                     self, instance: Any, owner: Union[type, None] = None

libs/core/langchain_core/_api/deprecation.py

@@ -225,7 +225,7 @@ def deprecated(
                 obj.__init__ = functools.wraps(obj.__init__)(  # type: ignore[misc]
                     warn_if_direct_instance
                 )
-                return cast("T", obj)
+                return obj
 
         elif isinstance(obj, FieldInfoV1):
             wrapped = None

libs/core/langchain_core/callbacks/base.py

@@ -7,6 +7,8 @@ from typing import TYPE_CHECKING, Any, Optional, Union
 
 from typing_extensions import Self
 
+from langchain_core.v1.messages import AIMessage, AIMessageChunk, MessageV1
+
 if TYPE_CHECKING:
     from collections.abc import Sequence
     from uuid import UUID
@@ -66,7 +68,9 @@ class LLMManagerMixin:
         self,
         token: str,
         *,
-        chunk: Optional[Union[GenerationChunk, ChatGenerationChunk]] = None,
+        chunk: Optional[
+            Union[GenerationChunk, ChatGenerationChunk, AIMessageChunk]
+        ] = None,
         run_id: UUID,
         parent_run_id: Optional[UUID] = None,
         **kwargs: Any,
@@ -75,8 +79,8 @@ class LLMManagerMixin:
 
         Args:
             token (str): The new token.
-            chunk (GenerationChunk | ChatGenerationChunk): The new generated chunk,
-              containing content and other information.
+            chunk (GenerationChunk | ChatGenerationChunk | AIMessageChunk): The new
+              generated chunk, containing content and other information.
             run_id (UUID): The run ID. This is the ID of the current run.
             parent_run_id (UUID): The parent run ID. This is the ID of the parent run.
             kwargs (Any): Additional keyword arguments.
@@ -84,7 +88,7 @@ class LLMManagerMixin:
 
     def on_llm_end(
         self,
-        response: LLMResult,
+        response: Union[LLMResult, AIMessage],
         *,
         run_id: UUID,
         parent_run_id: Optional[UUID] = None,
@@ -93,7 +97,7 @@ class LLMManagerMixin:
         """Run when LLM ends running.
 
         Args:
-            response (LLMResult): The response which was generated.
+            response (LLMResult | AIMessage): The response which was generated.
             run_id (UUID): The run ID. This is the ID of the current run.
             parent_run_id (UUID): The parent run ID. This is the ID of the parent run.
             kwargs (Any): Additional keyword arguments.
@@ -261,7 +265,7 @@ class CallbackManagerMixin:
     def on_chat_model_start(
         self,
         serialized: dict[str, Any],
-        messages: list[list[BaseMessage]],
+        messages: Union[list[list[BaseMessage]], list[MessageV1]],
         *,
         run_id: UUID,
         parent_run_id: Optional[UUID] = None,
@@ -439,6 +443,9 @@ class BaseCallbackHandler(
     run_inline: bool = False
     """Whether to run the callback inline."""
 
+    accepts_new_messages: bool = False
+    """Whether the callback accepts new message format."""
+
     @property
     def ignore_llm(self) -> bool:
         """Whether to ignore LLM callbacks."""
@@ -509,7 +516,7 @@ class AsyncCallbackHandler(BaseCallbackHandler):
     async def on_chat_model_start(
         self,
         serialized: dict[str, Any],
-        messages: list[list[BaseMessage]],
+        messages: Union[list[list[BaseMessage]], list[MessageV1]],
         *,
         run_id: UUID,
         parent_run_id: Optional[UUID] = None,
@@ -540,7 +547,9 @@ class AsyncCallbackHandler(BaseCallbackHandler):
         self,
         token: str,
         *,
-        chunk: Optional[Union[GenerationChunk, ChatGenerationChunk]] = None,
+        chunk: Optional[
+            Union[GenerationChunk, ChatGenerationChunk, AIMessageChunk]
+        ] = None,
         run_id: UUID,
         parent_run_id: Optional[UUID] = None,
         tags: Optional[list[str]] = None,
@@ -550,8 +559,8 @@ class AsyncCallbackHandler(BaseCallbackHandler):
 
         Args:
             token (str): The new token.
-            chunk (GenerationChunk | ChatGenerationChunk): The new generated chunk,
-              containing content and other information.
+            chunk (GenerationChunk | ChatGenerationChunk | AIMessageChunk): The new
+              generated chunk, containing content and other information.
             run_id (UUID): The run ID. This is the ID of the current run.
             parent_run_id (UUID): The parent run ID. This is the ID of the parent run.
             tags (Optional[list[str]]): The tags.
@@ -560,7 +569,7 @@ class AsyncCallbackHandler(BaseCallbackHandler):
 
     async def on_llm_end(
         self,
-        response: LLMResult,
+        response: Union[LLMResult, AIMessage],
         *,
         run_id: UUID,
         parent_run_id: Optional[UUID] = None,
@@ -570,7 +579,7 @@ class AsyncCallbackHandler(BaseCallbackHandler):
         """Run when LLM ends running.
 
         Args:
-            response (LLMResult): The response which was generated.
+            response (LLMResult | AIMessage): The response which was generated.
             run_id (UUID): The run ID. This is the ID of the current run.
             parent_run_id (UUID): The parent run ID. This is the ID of the parent run.
             tags (Optional[list[str]]): The tags.
@@ -594,8 +603,8 @@ class AsyncCallbackHandler(BaseCallbackHandler):
             parent_run_id: The parent run ID. This is the ID of the parent run.
             tags: The tags.
             kwargs (Any): Additional keyword arguments.
-                - response (LLMResult): The response which was generated before
-                    the error occurred.
+                - response (LLMResult | AIMessage): The response which was generated
+                    before the error occurred.
         """
 
     async def on_chain_start(

libs/core/langchain_core/callbacks/file.py

@@ -49,7 +49,7 @@ class FileCallbackHandler(BaseCallbackHandler):
         mode: The file open mode. Defaults to ``'a'`` (append).
         color: Default color for text output. Defaults to ``None``.
 
-    Note:
+    .. note::
         When not used as a context manager, a deprecation warning will be issued
         on first use. The file will be opened immediately in ``__init__`` and closed
         in ``__del__`` or when ``close()`` is called explicitly.
@@ -65,6 +65,7 @@ class FileCallbackHandler(BaseCallbackHandler):
             filename: Path to the output file.
             mode: File open mode (e.g., ``'w'``, ``'a'``, ``'x'``). Defaults to ``'a'``.
             color: Default text color for output. Defaults to ``None``.
+
         """
         self.filename = filename
         self.mode = mode
@@ -82,9 +83,10 @@ class FileCallbackHandler(BaseCallbackHandler):
         Returns:
             The FileCallbackHandler instance.
 
-        Note:
+        .. note::
             The file is already opened in ``__init__``, so this just marks that
             the handler is being used as a context manager.
+
         """
         self._file_opened_in_context = True
         return self
@@ -101,6 +103,7 @@ class FileCallbackHandler(BaseCallbackHandler):
             exc_type: Exception type if an exception occurred.
             exc_val: Exception value if an exception occurred.
             exc_tb: Exception traceback if an exception occurred.
+
         """
         self.close()
 
@@ -113,6 +116,7 @@ class FileCallbackHandler(BaseCallbackHandler):
 
         This method is safe to call multiple times and will only close
         the file if it's currently open.
+
         """
         if hasattr(self, "file") and self.file and not self.file.closed:
             self.file.close()
@@ -133,6 +137,7 @@ class FileCallbackHandler(BaseCallbackHandler):
 
         Raises:
             RuntimeError: If the file is closed or not available.
+
         """
         global _GLOBAL_DEPRECATION_WARNED  # noqa: PLW0603
         if not self._file_opened_in_context and not _GLOBAL_DEPRECATION_WARNED:
@@ -163,6 +168,7 @@ class FileCallbackHandler(BaseCallbackHandler):
             serialized: The serialized chain information.
             inputs: The inputs to the chain.
             **kwargs: Additional keyword arguments that may contain ``'name'``.
+
         """
         name = (
             kwargs.get("name")
@@ -178,6 +184,7 @@ class FileCallbackHandler(BaseCallbackHandler):
         Args:
             outputs: The outputs of the chain.
             **kwargs: Additional keyword arguments.
+
         """
         self._write("\n> Finished chain.", end="\n")
 
@@ -192,6 +199,7 @@ class FileCallbackHandler(BaseCallbackHandler):
             color: Color override for this specific output. If ``None``, uses
                 ``self.color``.
             **kwargs: Additional keyword arguments.
+
         """
         self._write(action.log, color=color or self.color)
 
@@ -213,6 +221,7 @@ class FileCallbackHandler(BaseCallbackHandler):
             observation_prefix: Optional prefix to write before the output.
             llm_prefix: Optional prefix to write after the output.
             **kwargs: Additional keyword arguments.
+
         """
         if observation_prefix is not None:
             self._write(f"\n{observation_prefix}")
@@ -232,6 +241,7 @@ class FileCallbackHandler(BaseCallbackHandler):
                 ``self.color``.
             end: String appended after the text. Defaults to ``""``.
             **kwargs: Additional keyword arguments.
+
         """
         self._write(text, color=color or self.color, end=end)
 
@@ -246,5 +256,6 @@ class FileCallbackHandler(BaseCallbackHandler):
             color: Color override for this specific output. If ``None``, uses
                 ``self.color``.
             **kwargs: Additional keyword arguments.
+
         """
         self._write(finish.log, color=color or self.color, end="\n")

libs/core/langchain_core/callbacks/manager.py

@@ -11,15 +11,7 @@ from abc import ABC, abstractmethod
 from concurrent.futures import ThreadPoolExecutor
 from contextlib import asynccontextmanager, contextmanager
 from contextvars import copy_context
-from typing import (
-    TYPE_CHECKING,
-    Any,
-    Callable,
-    Optional,
-    TypeVar,
-    Union,
-    cast,
-)
+from typing import TYPE_CHECKING, Any, Callable, Optional, TypeVar, Union, cast
 from uuid import UUID
 
 from langsmith.run_helpers import get_tracing_context
@@ -37,8 +29,16 @@ from langchain_core.callbacks.base import (
 )
 from langchain_core.callbacks.stdout import StdOutCallbackHandler
 from langchain_core.messages import BaseMessage, get_buffer_string
+from langchain_core.messages.utils import convert_from_v1_message
+from langchain_core.outputs import ChatGeneration, ChatGenerationChunk, LLMResult
 from langchain_core.tracers.schemas import Run
 from langchain_core.utils.env import env_var_is_set
+from langchain_core.v1.messages import (
+    AIMessage,
+    AIMessageChunk,
+    MessageV1,
+    MessageV1Types,
+)
 
 if TYPE_CHECKING:
     from collections.abc import AsyncGenerator, Coroutine, Generator, Sequence
@@ -47,7 +47,7 @@ if TYPE_CHECKING:
 
     from langchain_core.agents import AgentAction, AgentFinish
     from langchain_core.documents import Document
-    from langchain_core.outputs import ChatGenerationChunk, GenerationChunk, LLMResult
+    from langchain_core.outputs import GenerationChunk
     from langchain_core.runnables.config import RunnableConfig
 
 logger = logging.getLogger(__name__)
@@ -92,7 +92,8 @@ def trace_as_chain_group(
         metadata (dict[str, Any], optional): The metadata to apply to all runs.
             Defaults to None.
 
-    Note: must have LANGCHAIN_TRACING_V2 env var set to true to see the trace in LangSmith.
+    .. note:
+        Must have ``LANGCHAIN_TRACING_V2`` env var set to true to see the trace in LangSmith.
 
     Returns:
         CallbackManagerForChainGroup: The callback manager for the chain group.
@@ -177,7 +178,8 @@ async def atrace_as_chain_group(
     Returns:
         AsyncCallbackManager: The async callback manager for the chain group.
 
-    Note: must have LANGCHAIN_TRACING_V2 env var set to true to see the trace in LangSmith.
+    .. note:
+        Must have ``LANGCHAIN_TRACING_V2`` env var set to true to see the trace in LangSmith.
 
     Example:
         .. code-block:: python
@@ -234,6 +236,7 @@ def shielded(func: Func) -> Func:
 
     Returns:
         Callable: The shielded function
+
     """
 
     @functools.wraps(func)
@@ -243,6 +246,46 @@ def shielded(func: Func) -> Func:
     return cast("Func", wrapped)
 
 
+def _convert_llm_events(
+    event_name: str, args: tuple[Any, ...], kwargs: dict[str, Any]
+) -> tuple[tuple[Any, ...], dict[str, Any]]:
+    args_list = list(args)
+    if (
+        event_name == "on_chat_model_start"
+        and isinstance(args_list[1], list)
+        and args_list[1]
+        and isinstance(args_list[1][0], MessageV1Types)
+    ):
+        batch = [
+            convert_from_v1_message(item)
+            for item in args_list[1]
+            if isinstance(item, MessageV1Types)
+        ]
+        args_list[1] = [batch]
+    elif (
+        event_name == "on_llm_new_token"
+        and "chunk" in kwargs
+        and isinstance(kwargs["chunk"], MessageV1Types)
+    ):
+        chunk = kwargs["chunk"]
+        kwargs["chunk"] = ChatGenerationChunk(text=chunk.text, message=chunk)
+    elif event_name == "on_llm_end" and isinstance(args_list[0], MessageV1Types):
+        args_list[0] = LLMResult(
+            generations=[
+                [
+                    ChatGeneration(
+                        text=args_list[0].text,
+                        message=convert_from_v1_message(args_list[0]),
+                    )
+                ]
+            ]
+        )
+    else:
+        pass
+
+    return tuple(args_list), kwargs
+
+
 def handle_event(
     handlers: list[BaseCallbackHandler],
     event_name: str,
@@ -252,15 +295,17 @@ def handle_event(
 ) -> None:
     """Generic event handler for CallbackManager.
 
-    Note: This function is used by LangServe to handle events.
+    .. note::
+        This function is used by ``LangServe`` to handle events.
 
     Args:
         handlers: The list of handlers that will handle the event.
-        event_name: The name of the event (e.g., "on_llm_start").
+        event_name: The name of the event (e.g., ``'on_llm_start'``).
         ignore_condition_name: Name of the attribute defined on handler
             that if True will cause the handler to be skipped for the given event.
         *args: The arguments to pass to the event handler.
         **kwargs: The keyword arguments to pass to the event handler
+
     """
     coros: list[Coroutine[Any, Any, Any]] = []
 
@@ -271,6 +316,8 @@ def handle_event(
                 if ignore_condition_name is None or not getattr(
                     handler, ignore_condition_name
                 ):
+                    if not handler.accepts_new_messages:
+                        args, kwargs = _convert_llm_events(event_name, args, kwargs)
                     event = getattr(handler, event_name)(*args, **kwargs)
                     if asyncio.iscoroutine(event):
                         coros.append(event)
@@ -365,6 +412,8 @@ async def _ahandle_event_for_handler(
 ) -> None:
     try:
         if ignore_condition_name is None or not getattr(handler, ignore_condition_name):
+            if not handler.accepts_new_messages:
+                args, kwargs = _convert_llm_events(event_name, args, kwargs)
             event = getattr(handler, event_name)
             if asyncio.iscoroutinefunction(event):
                 await event(*args, **kwargs)
@@ -415,17 +464,19 @@ async def ahandle_event(
     *args: Any,
     **kwargs: Any,
 ) -> None:
-    """Async generic event handler for AsyncCallbackManager.
+    """Async generic event handler for ``AsyncCallbackManager``.
 
-    Note: This function is used by LangServe to handle events.
+    .. note::
+        This function is used by ``LangServe`` to handle events.
 
     Args:
         handlers: The list of handlers that will handle the event.
-        event_name: The name of the event (e.g., "on_llm_start").
+        event_name: The name of the event (e.g., ``'on_llm_start'``).
         ignore_condition_name: Name of the attribute defined on handler
             that if True will cause the handler to be skipped for the given event.
         *args: The arguments to pass to the event handler.
         **kwargs: The keyword arguments to pass to the event handler.
+
     """
     for handler in [h for h in handlers if h.run_inline]:
         await _ahandle_event_for_handler(
@@ -477,6 +528,7 @@ class BaseRunManager(RunManagerMixin):
                 Defaults to None.
             inheritable_metadata (Optional[dict[str, Any]]): The inheritable metadata.
                 Defaults to None.
+
         """
         self.run_id = run_id
         self.handlers = handlers
@@ -493,6 +545,7 @@ class BaseRunManager(RunManagerMixin):
 
         Returns:
             BaseRunManager: The noop manager.
+
         """
         return cls(
             run_id=uuid.uuid4(),
@@ -545,6 +598,7 @@ class RunManager(BaseRunManager):
         Args:
             retry_state (RetryCallState): The retry state.
             **kwargs (Any): Additional keyword arguments.
+
         """
         if not self.handlers:
             return
@@ -572,6 +626,7 @@ class ParentRunManager(RunManager):
 
         Returns:
             CallbackManager: The child callback manager.
+
         """
         manager = CallbackManager(handlers=[], parent_run_id=self.run_id)
         manager.set_handlers(self.inheritable_handlers)
@@ -591,6 +646,7 @@ class AsyncRunManager(BaseRunManager, ABC):
 
         Returns:
             RunManager: The sync RunManager.
+
         """
 
     async def on_text(
@@ -606,6 +662,7 @@ class AsyncRunManager(BaseRunManager, ABC):
 
         Returns:
             Any: The result of the callback.
+
         """
         if not self.handlers:
             return
@@ -630,6 +687,7 @@ class AsyncRunManager(BaseRunManager, ABC):
         Args:
             retry_state (RetryCallState): The retry state.
             **kwargs (Any): Additional keyword arguments.
+
         """
         if not self.handlers:
             return
@@ -657,6 +715,7 @@ class AsyncParentRunManager(AsyncRunManager):
 
         Returns:
             AsyncCallbackManager: The child callback manager.
+
         """
         manager = AsyncCallbackManager(handlers=[], parent_run_id=self.run_id)
         manager.set_handlers(self.inheritable_handlers)
@@ -674,7 +733,9 @@ class CallbackManagerForLLMRun(RunManager, LLMManagerMixin):
         self,
         token: str,
         *,
-        chunk: Optional[Union[GenerationChunk, ChatGenerationChunk]] = None,
+        chunk: Optional[
+            Union[GenerationChunk, ChatGenerationChunk, AIMessageChunk]
+        ] = None,
         **kwargs: Any,
     ) -> None:
         """Run when LLM generates a new token.
@@ -684,6 +745,7 @@ class CallbackManagerForLLMRun(RunManager, LLMManagerMixin):
             chunk (Optional[Union[GenerationChunk, ChatGenerationChunk]], optional):
                 The chunk. Defaults to None.
             **kwargs (Any): Additional keyword arguments.
+
         """
         if not self.handlers:
             return
@@ -699,12 +761,13 @@ class CallbackManagerForLLMRun(RunManager, LLMManagerMixin):
             **kwargs,
         )
 
-    def on_llm_end(self, response: LLMResult, **kwargs: Any) -> None:
+    def on_llm_end(self, response: Union[LLMResult, AIMessage], **kwargs: Any) -> None:
         """Run when LLM ends running.
 
         Args:
-            response (LLMResult): The LLM result.
+            response (LLMResult | AIMessage): The LLM result.
             **kwargs (Any): Additional keyword arguments.
+
         """
         if not self.handlers:
             return
@@ -729,8 +792,9 @@ class CallbackManagerForLLMRun(RunManager, LLMManagerMixin):
         Args:
             error (Exception or KeyboardInterrupt): The error.
             kwargs (Any): Additional keyword arguments.
-                - response (LLMResult): The response which was generated before
-                    the error occurred.
+                - response (LLMResult | AIMessage): The response which was generated
+                  before the error occurred.
+
         """
         if not self.handlers:
             return
@@ -754,6 +818,7 @@ class AsyncCallbackManagerForLLMRun(AsyncRunManager, LLMManagerMixin):
 
         Returns:
             CallbackManagerForLLMRun: The sync RunManager.
+
         """
         return CallbackManagerForLLMRun(
             run_id=self.run_id,
@@ -770,7 +835,9 @@ class AsyncCallbackManagerForLLMRun(AsyncRunManager, LLMManagerMixin):
         self,
         token: str,
         *,
-        chunk: Optional[Union[GenerationChunk, ChatGenerationChunk]] = None,
+        chunk: Optional[
+            Union[GenerationChunk, ChatGenerationChunk, AIMessageChunk]
+        ] = None,
         **kwargs: Any,
     ) -> None:
         """Run when LLM generates a new token.
@@ -780,6 +847,7 @@ class AsyncCallbackManagerForLLMRun(AsyncRunManager, LLMManagerMixin):
             chunk (Optional[Union[GenerationChunk, ChatGenerationChunk]], optional):
                 The chunk. Defaults to None.
             **kwargs (Any): Additional keyword arguments.
+
         """
         if not self.handlers:
             return
@@ -796,12 +864,15 @@ class AsyncCallbackManagerForLLMRun(AsyncRunManager, LLMManagerMixin):
         )
 
     @shielded
-    async def on_llm_end(self, response: LLMResult, **kwargs: Any) -> None:
+    async def on_llm_end(
+        self, response: Union[LLMResult, AIMessage], **kwargs: Any
+    ) -> None:
         """Run when LLM ends running.
 
         Args:
-            response (LLMResult): The LLM result.
+            response (LLMResult | AIMessage): The LLM result.
             **kwargs (Any): Additional keyword arguments.
+
         """
         if not self.handlers:
             return
@@ -827,10 +898,8 @@ class AsyncCallbackManagerForLLMRun(AsyncRunManager, LLMManagerMixin):
         Args:
             error (Exception or KeyboardInterrupt): The error.
             kwargs (Any): Additional keyword arguments.
-                - response (LLMResult): The response which was generated before
-                    the error occurred.
-
-
+                - response (LLMResult | AIMessage): The response which was generated
+                  before the error occurred.
 
         """
         if not self.handlers:
@@ -856,6 +925,7 @@ class CallbackManagerForChainRun(ParentRunManager, ChainManagerMixin):
         Args:
             outputs (Union[dict[str, Any], Any]): The outputs of the chain.
             **kwargs (Any): Additional keyword arguments.
+
         """
         if not self.handlers:
             return
@@ -880,6 +950,7 @@ class CallbackManagerForChainRun(ParentRunManager, ChainManagerMixin):
         Args:
             error (Exception or KeyboardInterrupt): The error.
             **kwargs (Any): Additional keyword arguments.
+
         """
         if not self.handlers:
             return
@@ -903,6 +974,7 @@ class CallbackManagerForChainRun(ParentRunManager, ChainManagerMixin):
 
         Returns:
             Any: The result of the callback.
+
         """
         if not self.handlers:
             return
@@ -926,6 +998,7 @@ class CallbackManagerForChainRun(ParentRunManager, ChainManagerMixin):
 
         Returns:
             Any: The result of the callback.
+
         """
         if not self.handlers:
             return
@@ -970,6 +1043,7 @@ class AsyncCallbackManagerForChainRun(AsyncParentRunManager, ChainManagerMixin):
         Args:
             outputs (Union[dict[str, Any], Any]): The outputs of the chain.
             **kwargs (Any): Additional keyword arguments.
+
         """
         if not self.handlers:
             return
@@ -995,6 +1069,7 @@ class AsyncCallbackManagerForChainRun(AsyncParentRunManager, ChainManagerMixin):
         Args:
             error (Exception or KeyboardInterrupt): The error.
             **kwargs (Any): Additional keyword arguments.
+
         """
         if not self.handlers:
             return
@@ -1018,6 +1093,7 @@ class AsyncCallbackManagerForChainRun(AsyncParentRunManager, ChainManagerMixin):
 
         Returns:
             Any: The result of the callback.
+
         """
         if not self.handlers:
             return
@@ -1041,6 +1117,7 @@ class AsyncCallbackManagerForChainRun(AsyncParentRunManager, ChainManagerMixin):
 
         Returns:
             Any: The result of the callback.
+
         """
         if not self.handlers:
             return
@@ -1069,6 +1146,7 @@ class CallbackManagerForToolRun(ParentRunManager, ToolManagerMixin):
         Args:
             output (Any): The output of the tool.
             **kwargs (Any): The keyword arguments to pass to the event handler
+
         """
         if not self.handlers:
             return
@@ -1093,6 +1171,7 @@ class CallbackManagerForToolRun(ParentRunManager, ToolManagerMixin):
         Args:
             error (Exception or KeyboardInterrupt): The error.
             **kwargs (Any): Additional keyword arguments.
+
         """
         if not self.handlers:
             return
@@ -1134,6 +1213,7 @@ class AsyncCallbackManagerForToolRun(AsyncParentRunManager, ToolManagerMixin):
         Args:
             output (Any): The output of the tool.
             **kwargs (Any): Additional keyword arguments.
+
         """
         if not self.handlers:
             return
@@ -1158,6 +1238,7 @@ class AsyncCallbackManagerForToolRun(AsyncParentRunManager, ToolManagerMixin):
         Args:
             error (Exception or KeyboardInterrupt): The error.
             **kwargs (Any): Additional keyword arguments.
+
         """
         if not self.handlers:
             return
@@ -1186,6 +1267,7 @@ class CallbackManagerForRetrieverRun(ParentRunManager, RetrieverManagerMixin):
         Args:
             documents (Sequence[Document]): The retrieved documents.
             **kwargs (Any): Additional keyword arguments.
+
         """
         if not self.handlers:
             return
@@ -1210,6 +1292,7 @@ class CallbackManagerForRetrieverRun(ParentRunManager, RetrieverManagerMixin):
         Args:
             error (BaseException): The error.
             **kwargs (Any): Additional keyword arguments.
+
         """
         if not self.handlers:
             return
@@ -1236,6 +1319,7 @@ class AsyncCallbackManagerForRetrieverRun(
 
         Returns:
             CallbackManagerForRetrieverRun: The sync RunManager.
+
         """
         return CallbackManagerForRetrieverRun(
             run_id=self.run_id,
@@ -1257,6 +1341,7 @@ class AsyncCallbackManagerForRetrieverRun(
         Args:
             documents (Sequence[Document]): The retrieved documents.
             **kwargs (Any): Additional keyword arguments.
+
         """
         if not self.handlers:
             return
@@ -1282,6 +1367,7 @@ class AsyncCallbackManagerForRetrieverRun(
         Args:
             error (BaseException): The error.
             **kwargs (Any): Additional keyword arguments.
+
         """
         if not self.handlers:
             return
@@ -1318,6 +1404,7 @@ class CallbackManager(BaseCallbackManager):
         Returns:
             list[CallbackManagerForLLMRun]: A callback manager for each
                 prompt as an LLM run.
+
         """
         managers = []
         for i, prompt in enumerate(prompts):
@@ -1354,7 +1441,7 @@ class CallbackManager(BaseCallbackManager):
     def on_chat_model_start(
         self,
         serialized: dict[str, Any],
-        messages: list[list[BaseMessage]],
+        messages: Union[list[list[BaseMessage]], list[MessageV1]],
         run_id: Optional[UUID] = None,
         **kwargs: Any,
     ) -> list[CallbackManagerForLLMRun]:
@@ -1362,14 +1449,41 @@ class CallbackManager(BaseCallbackManager):
 
         Args:
             serialized (dict[str, Any]): The serialized LLM.
-            messages (list[list[BaseMessage]]): The list of messages.
+            messages (list[list[BaseMessage | MessageV1]]): The list of messages.
             run_id (UUID, optional): The ID of the run. Defaults to None.
             **kwargs (Any): Additional keyword arguments.
 
         Returns:
             list[CallbackManagerForLLMRun]: A callback manager for each
                 list of messages as an LLM run.
+
         """
+        if messages and isinstance(messages[0], MessageV1Types):
+            run_id_ = run_id if run_id is not None else uuid.uuid4()
+            handle_event(
+                self.handlers,
+                "on_chat_model_start",
+                "ignore_chat_model",
+                serialized,
+                messages,
+                run_id=run_id_,
+                parent_run_id=self.parent_run_id,
+                tags=self.tags,
+                metadata=self.metadata,
+                **kwargs,
+            )
+            return [
+                CallbackManagerForLLMRun(
+                    run_id=run_id_,
+                    handlers=self.handlers,
+                    inheritable_handlers=self.inheritable_handlers,
+                    parent_run_id=self.parent_run_id,
+                    tags=self.tags,
+                    inheritable_tags=self.inheritable_tags,
+                    metadata=self.metadata,
+                    inheritable_metadata=self.inheritable_metadata,
+                )
+            ]
         managers = []
         for message_list in messages:
             if run_id is not None:
@@ -1422,6 +1536,7 @@ class CallbackManager(BaseCallbackManager):
 
         Returns:
             CallbackManagerForChainRun: The callback manager for the chain run.
+
         """
         if run_id is None:
             run_id = uuid.uuid4()
@@ -1476,6 +1591,7 @@ class CallbackManager(BaseCallbackManager):
 
         Returns:
             CallbackManagerForToolRun: The callback manager for the tool run.
+
         """
         if run_id is None:
             run_id = uuid.uuid4()
@@ -1522,6 +1638,7 @@ class CallbackManager(BaseCallbackManager):
             run_id (UUID, optional): The ID of the run. Defaults to None.
             parent_run_id (UUID, optional): The ID of the parent run. Defaults to None.
             **kwargs (Any): Additional keyword arguments.
+
         """
         if run_id is None:
             run_id = uuid.uuid4()
@@ -1569,6 +1686,7 @@ class CallbackManager(BaseCallbackManager):
             run_id: The ID of the run. Defaults to None.
 
         .. versionadded:: 0.2.14
+
         """
         if not self.handlers:
             return
@@ -1623,6 +1741,7 @@ class CallbackManager(BaseCallbackManager):
 
         Returns:
             CallbackManager: The configured callback manager.
+
         """
         return _configure(
             cls,
@@ -1657,6 +1776,7 @@ class CallbackManagerForChainGroup(CallbackManager):
             parent_run_id (Optional[UUID]): The ID of the parent run. Defaults to None.
             parent_run_manager (CallbackManagerForChainRun): The parent run manager.
             **kwargs (Any): Additional keyword arguments.
+
         """
         super().__init__(
             handlers,
@@ -1745,6 +1865,7 @@ class CallbackManagerForChainGroup(CallbackManager):
         Args:
             outputs (Union[dict[str, Any], Any]): The outputs of the chain.
             **kwargs (Any): Additional keyword arguments.
+
         """
         self.ended = True
         return self.parent_run_manager.on_chain_end(outputs, **kwargs)
@@ -1759,6 +1880,7 @@ class CallbackManagerForChainGroup(CallbackManager):
         Args:
             error (Exception or KeyboardInterrupt): The error.
             **kwargs (Any): Additional keyword arguments.
+
         """
         self.ended = True
         return self.parent_run_manager.on_chain_error(error, **kwargs)
@@ -1864,7 +1986,7 @@ class AsyncCallbackManager(BaseCallbackManager):
     async def on_chat_model_start(
         self,
         serialized: dict[str, Any],
-        messages: list[list[BaseMessage]],
+        messages: Union[list[list[BaseMessage]], list[MessageV1]],
         run_id: Optional[UUID] = None,
         **kwargs: Any,
     ) -> list[AsyncCallbackManagerForLLMRun]:
@@ -1872,7 +1994,7 @@ class AsyncCallbackManager(BaseCallbackManager):
 
         Args:
             serialized (dict[str, Any]): The serialized LLM.
-            messages (list[list[BaseMessage]]): The list of messages.
+            messages (list[list[BaseMessage | MessageV1]]): The list of messages.
             run_id (UUID, optional): The ID of the run. Defaults to None.
             **kwargs (Any): Additional keyword arguments.
 
@@ -1881,10 +2003,51 @@ class AsyncCallbackManager(BaseCallbackManager):
                 async callback managers, one for each LLM Run
                 corresponding to each inner  message list.
         """
+        if messages and isinstance(messages[0], MessageV1Types):
+            run_id_ = run_id if run_id is not None else uuid.uuid4()
+            inline_tasks = []
+            non_inline_tasks = []
+            for handler in self.handlers:
+                task = ahandle_event(
+                    [handler],
+                    "on_chat_model_start",
+                    "ignore_chat_model",
+                    serialized,
+                    messages,
+                    run_id=run_id_,
+                    parent_run_id=self.parent_run_id,
+                    tags=self.tags,
+                    metadata=self.metadata,
+                    **kwargs,
+                )
+                if handler.run_inline:
+                    inline_tasks.append(task)
+                else:
+                    non_inline_tasks.append(task)
+            managers = [
+                AsyncCallbackManagerForLLMRun(
+                    run_id=run_id_,
+                    handlers=self.handlers,
+                    inheritable_handlers=self.inheritable_handlers,
+                    parent_run_id=self.parent_run_id,
+                    tags=self.tags,
+                    inheritable_tags=self.inheritable_tags,
+                    metadata=self.metadata,
+                    inheritable_metadata=self.inheritable_metadata,
+                )
+            ]
+            # Run inline tasks sequentially
+            for task in inline_tasks:
+                await task
+
+            # Run non-inline tasks concurrently
+            if non_inline_tasks:
+                await asyncio.gather(*non_inline_tasks)
+
+            return managers
         inline_tasks = []
         non_inline_tasks = []
         managers = []
-
         for message_list in messages:
             if run_id is not None:
                 run_id_ = run_id

libs/core/langchain_core/callbacks/streaming_stdout.py

@@ -3,7 +3,7 @@
 from __future__ import annotations
 
 import sys
-from typing import TYPE_CHECKING, Any
+from typing import TYPE_CHECKING, Any, Union
 
 from typing_extensions import override
 
@@ -13,6 +13,7 @@ if TYPE_CHECKING:
     from langchain_core.agents import AgentAction, AgentFinish
     from langchain_core.messages import BaseMessage
     from langchain_core.outputs import LLMResult
+    from langchain_core.v1.messages import AIMessage, MessageV1
 
 
 class StreamingStdOutCallbackHandler(BaseCallbackHandler):
@@ -32,7 +33,7 @@ class StreamingStdOutCallbackHandler(BaseCallbackHandler):
     def on_chat_model_start(
         self,
         serialized: dict[str, Any],
-        messages: list[list[BaseMessage]],
+        messages: Union[list[list[BaseMessage]], list[MessageV1]],
         **kwargs: Any,
     ) -> None:
         """Run when LLM starts running.
@@ -54,7 +55,7 @@ class StreamingStdOutCallbackHandler(BaseCallbackHandler):
         sys.stdout.write(token)
         sys.stdout.flush()
 
-    def on_llm_end(self, response: LLMResult, **kwargs: Any) -> None:
+    def on_llm_end(self, response: Union[LLMResult, AIMessage], **kwargs: Any) -> None:
         """Run when LLM ends running.
 
         Args:

libs/core/langchain_core/callbacks/usage.py

@@ -4,14 +4,16 @@ import threading
 from collections.abc import Generator
 from contextlib import contextmanager
 from contextvars import ContextVar
-from typing import Any, Optional
+from typing import Any, Optional, Union
 
 from typing_extensions import override
 
 from langchain_core.callbacks import BaseCallbackHandler
 from langchain_core.messages import AIMessage
 from langchain_core.messages.ai import UsageMetadata, add_usage
+from langchain_core.messages.utils import convert_from_v1_message
 from langchain_core.outputs import ChatGeneration, LLMResult
+from langchain_core.v1.messages import AIMessage as AIMessageV1
 
 
 class UsageMetadataCallbackHandler(BaseCallbackHandler):
@@ -58,9 +60,17 @@ class UsageMetadataCallbackHandler(BaseCallbackHandler):
         return str(self.usage_metadata)
 
     @override
-    def on_llm_end(self, response: LLMResult, **kwargs: Any) -> None:
+    def on_llm_end(
+        self, response: Union[LLMResult, AIMessageV1], **kwargs: Any
+    ) -> None:
         """Collect token usage."""
         # Check for usage_metadata (langchain-core >= 0.2.2)
+        if isinstance(response, AIMessageV1):
+            response = LLMResult(
+                generations=[
+                    [ChatGeneration(message=convert_from_v1_message(response))]
+                ]
+            )
         try:
             generation = response.generations[0][0]
         except IndexError:

libs/core/langchain_core/chat_history.py

@@ -117,9 +117,9 @@ class BaseChatMessageHistory(ABC):
     def add_user_message(self, message: Union[HumanMessage, str]) -> None:
         """Convenience method for adding a human message string to the store.
 
-        Please note that this is a convenience method. Code should favor the
-        bulk add_messages interface instead to save on round-trips to the underlying
-        persistence layer.
+        .. note::
+            This is a convenience method. Code should favor the bulk ``add_messages``
+            interface instead to save on round-trips to the persistence layer.
 
         This method may be deprecated in a future release.
 
@@ -134,9 +134,9 @@ class BaseChatMessageHistory(ABC):
     def add_ai_message(self, message: Union[AIMessage, str]) -> None:
         """Convenience method for adding an AI message string to the store.
 
-        Please note that this is a convenience method. Code should favor the bulk
-        add_messages interface instead to save on round-trips to the underlying
-        persistence layer.
+        .. note::
+            This is a convenience method. Code should favor the bulk ``add_messages``
+            interface instead to save on round-trips to the persistence layer.
 
         This method may be deprecated in a future release.
 

libs/core/langchain_core/documents/compressor.py

@@ -19,17 +19,18 @@ if TYPE_CHECKING:
 class BaseDocumentCompressor(BaseModel, ABC):
     """Base class for document compressors.
 
-    This abstraction is primarily used for
-    post-processing of retrieved documents.
+    This abstraction is primarily used for post-processing of retrieved documents.
 
     Documents matching a given query are first retrieved.
+
     Then the list of documents can be further processed.
 
-    For example, one could re-rank the retrieved documents
-    using an LLM.
+    For example, one could re-rank the retrieved documents using an LLM.
+
+    .. note::
+        Users should favor using a RunnableLambda instead of sub-classing from this
+        interface.
 
-    **Note** users should favor using a RunnableLambda
-    instead of sub-classing from this interface.
     """
 
     @abstractmethod
@@ -48,6 +49,7 @@ class BaseDocumentCompressor(BaseModel, ABC):
 
         Returns:
             The compressed documents.
+
         """
 
     async def acompress_documents(
@@ -65,6 +67,7 @@ class BaseDocumentCompressor(BaseModel, ABC):
 
         Returns:
             The compressed documents.
+
         """
         return await run_in_executor(
             None, self.compress_documents, documents, query, callbacks

libs/core/langchain_core/indexing/base.py

@@ -488,8 +488,8 @@ class DeleteResponse(TypedDict, total=False):
     failed: Sequence[str]
     """The IDs that failed to be deleted.
 
-    Please note that deleting an ID that
-    does not exist is **NOT** considered a failure.
+    .. warning::
+        Deleting an ID that does not exist is **NOT** considered a failure.
     """
 
     num_failed: int

libs/core/langchain_core/language_models/_utils.py

@@ -1,8 +1,10 @@
+import copy
 import re
 from collections.abc import Sequence
 from typing import Optional
 
 from langchain_core.messages import BaseMessage
+from langchain_core.v1.messages import MessageV1
 
 
 def _is_openai_data_block(block: dict) -> bool:
@@ -138,3 +140,37 @@ def _normalize_messages(messages: Sequence[BaseMessage]) -> list[BaseMessage]:
         formatted_messages.append(formatted_message)
 
     return formatted_messages
+
+
+def _normalize_messages_v1(messages: Sequence[MessageV1]) -> list[MessageV1]:
+    """Extend support for message formats.
+
+    Chat models implement support for images in OpenAI Chat Completions format, as well
+    as other multimodal data as standard data blocks. This function extends support to
+    audio and file data in OpenAI Chat Completions format by converting them to standard
+    data blocks.
+    """
+    formatted_messages = []
+    for message in messages:
+        formatted_message = message
+        if isinstance(message.content, list):
+            for idx, block in enumerate(message.content):
+                if (
+                    isinstance(block, dict)
+                    # Subset to (PDF) files and audio, as most relevant chat models
+                    # support images in OAI format (and some may not yet support the
+                    # standard data block format)
+                    and block.get("type") in {"file", "input_audio"}
+                    and _is_openai_data_block(block)  # type: ignore[arg-type]
+                ):
+                    if formatted_message is message:
+                        formatted_message = copy.copy(message)
+                        # Also shallow-copy content
+                        formatted_message.content = list(formatted_message.content)
+
+                    formatted_message.content[idx] = (  # type: ignore[call-overload]
+                        _convert_openai_format_to_data_block(block)  # type: ignore[arg-type]
+                    )
+        formatted_messages.append(formatted_message)
+
+    return formatted_messages

libs/core/langchain_core/language_models/base.py

@@ -31,6 +31,7 @@ from langchain_core.messages import (
 from langchain_core.prompt_values import PromptValue
 from langchain_core.runnables import Runnable, RunnableSerializable
 from langchain_core.utils import get_pydantic_field_names
+from langchain_core.v1.messages import AIMessage as AIMessageV1
 
 if TYPE_CHECKING:
     from langchain_core.outputs import LLMResult
@@ -57,8 +58,8 @@ class LangSmithParams(TypedDict, total=False):
 def get_tokenizer() -> Any:
     """Get a GPT-2 tokenizer instance.
 
-    This function is cached to avoid re-loading the tokenizer
-    every time it is called.
+    This function is cached to avoid re-loading the tokenizer every time it is called.
+
     """
     try:
         from transformers import GPT2TokenizerFast  # type: ignore[import-not-found]
@@ -85,7 +86,9 @@ def _get_token_ids_default_method(text: str) -> list[int]:
 LanguageModelInput = Union[PromptValue, str, Sequence[MessageLikeRepresentation]]
 LanguageModelOutput = Union[BaseMessage, str]
 LanguageModelLike = Runnable[LanguageModelInput, LanguageModelOutput]
-LanguageModelOutputVar = TypeVar("LanguageModelOutputVar", BaseMessage, str)
+LanguageModelOutputVar = TypeVar(
+    "LanguageModelOutputVar", BaseMessage, str, AIMessageV1
+)
 
 
 def _get_verbosity() -> bool:
@@ -99,7 +102,8 @@ class BaseLanguageModel(
 ):
     """Abstract base class for interfacing with language models.
 
-    All language model wrappers inherited from BaseLanguageModel.
+    All language model wrappers inherited from ``BaseLanguageModel``.
+
     """
 
     cache: Union[BaseCache, bool, None] = Field(default=None, exclude=True)
@@ -108,9 +112,10 @@ class BaseLanguageModel(
     * If true, will use the global cache.
     * If false, will not use a cache
     * If None, will use the global cache if it's set, otherwise no cache.
-    * If instance of BaseCache, will use the provided cache.
+    * If instance of ``BaseCache``, will use the provided cache.
 
     Caching is not currently supported for streaming methods of models.
+
     """
     verbose: bool = Field(default_factory=_get_verbosity, exclude=True, repr=False)
     """Whether to print out response text."""
@@ -140,6 +145,7 @@ class BaseLanguageModel(
 
         Returns:
             The verbosity setting to use.
+
         """
         if verbose is None:
             return _get_verbosity()
@@ -195,7 +201,8 @@ class BaseLanguageModel(
 
         Returns:
             An LLMResult, which contains a list of candidate Generations for each input
-                prompt and additional model provider-specific output.
+            prompt and additional model provider-specific output.
+
         """
 
     @abstractmethod
@@ -229,8 +236,9 @@ class BaseLanguageModel(
                 to the model provider API call.
 
         Returns:
-            An LLMResult, which contains a list of candidate Generations for each input
-                prompt and additional model provider-specific output.
+            An ``LLMResult``, which contains a list of candidate Generations for each
+            input prompt and additional model provider-specific output.
+
         """
 
     def with_structured_output(
@@ -248,8 +256,8 @@ class BaseLanguageModel(
     ) -> str:
         """Pass a single string input to the model and return a string.
 
-         Use this method when passing in raw text. If you want to pass in specific
-            types of chat messages, use predict_messages.
+        Use this method when passing in raw text. If you want to pass in specific types
+        of chat messages, use predict_messages.
 
         Args:
             text: String input to pass to the model.
@@ -260,6 +268,7 @@ class BaseLanguageModel(
 
         Returns:
             Top model prediction as a string.
+
         """
 
     @deprecated("0.1.7", alternative="invoke", removal="1.0")
@@ -274,7 +283,7 @@ class BaseLanguageModel(
         """Pass a message sequence to the model and return a message.
 
         Use this method when passing in chat messages. If you want to pass in raw text,
-            use predict.
+        use predict.
 
         Args:
             messages: A sequence of chat messages corresponding to a single model input.
@@ -285,6 +294,7 @@ class BaseLanguageModel(
 
         Returns:
             Top model prediction as a message.
+
         """
 
     @deprecated("0.1.7", alternative="ainvoke", removal="1.0")
@@ -295,7 +305,7 @@ class BaseLanguageModel(
         """Asynchronously pass a string to the model and return a string.
 
         Use this method when calling pure text generation models and only the top
-            candidate generation is needed.
+        candidate generation is needed.
 
         Args:
             text: String input to pass to the model.
@@ -306,6 +316,7 @@ class BaseLanguageModel(
 
         Returns:
             Top model prediction as a string.
+
         """
 
     @deprecated("0.1.7", alternative="ainvoke", removal="1.0")
@@ -319,8 +330,8 @@ class BaseLanguageModel(
     ) -> BaseMessage:
         """Asynchronously pass messages to the model and return a message.
 
-        Use this method when calling chat models and only the top
-            candidate generation is needed.
+        Use this method when calling chat models and only the top candidate generation
+        is needed.
 
         Args:
             messages: A sequence of chat messages corresponding to a single model input.
@@ -331,6 +342,7 @@ class BaseLanguageModel(
 
         Returns:
             Top model prediction as a message.
+
         """
 
     @property
@@ -346,7 +358,8 @@ class BaseLanguageModel(
 
         Returns:
             A list of ids corresponding to the tokens in the text, in order they occur
-                in the text.
+            in the text.
+
         """
         if self.custom_get_token_ids is not None:
             return self.custom_get_token_ids(text)
@@ -362,6 +375,7 @@ class BaseLanguageModel(
 
         Returns:
             The integer number of tokens in the text.
+
         """
         return len(self.get_token_ids(text))
 
@@ -374,16 +388,18 @@ class BaseLanguageModel(
 
         Useful for checking if an input fits in a model's context window.
 
-        **Note**: the base implementation of get_num_tokens_from_messages ignores
-        tool schemas.
+        .. note::
+            The base implementation of ``get_num_tokens_from_messages`` ignores tool
+            schemas.
 
         Args:
             messages: The message inputs to tokenize.
-            tools: If provided, sequence of dict, BaseModel, function, or BaseTools
-                to be converted to tool schemas.
+            tools: If provided, sequence of dict, ``BaseModel``, function, or
+                ``BaseTools`` to be converted to tool schemas.
 
         Returns:
             The sum of the number of tokens across the messages.
+
         """
         if tools is not None:
             warnings.warn(
@@ -396,6 +412,7 @@ class BaseLanguageModel(
     def _all_required_field_names(cls) -> set:
         """DEPRECATED: Kept for backwards compatibility.
 
-        Use get_pydantic_field_names.
+        Use ``get_pydantic_field_names``.
+
         """
         return get_pydantic_field_names(cls)

libs/core/langchain_core/language_models/chat_models.py

@@ -97,17 +97,18 @@ def _generate_response_from_error(error: BaseException) -> list[ChatGeneration]:
 
 
 def _format_for_tracing(messages: list[BaseMessage]) -> list[BaseMessage]:
-    """Format messages for tracing in on_chat_model_start.
+    """Format messages for tracing in ``on_chat_model_start``.
 
     - Update image content blocks to OpenAI Chat Completions format (backward
     compatibility).
-    - Add "type" key to content blocks that have a single key.
+    - Add ``type`` key to content blocks that have a single key.
 
     Args:
         messages: List of messages to format.
 
     Returns:
         List of messages formatted for tracing.
+
     """
     messages_to_trace = []
     for message in messages:
@@ -153,10 +154,11 @@ def generate_from_stream(stream: Iterator[ChatGenerationChunk]) -> ChatResult:
     """Generate from a stream.
 
     Args:
-        stream: Iterator of ChatGenerationChunk.
+        stream: Iterator of ``ChatGenerationChunk``.
 
     Returns:
         ChatResult: Chat result.
+
     """
     generation = next(stream, None)
     if generation:
@@ -180,10 +182,11 @@ async def agenerate_from_stream(
     """Async generate from a stream.
 
     Args:
-        stream: Iterator of ChatGenerationChunk.
+        stream: Iterator of ``ChatGenerationChunk``.
 
     Returns:
         ChatResult: Chat result.
+
     """
     chunks = [chunk async for chunk in stream]
     return await run_in_executor(None, generate_from_stream, iter(chunks))
@@ -311,15 +314,16 @@ class BaseChatModel(BaseLanguageModel[BaseMessage], ABC):
       provided. This offers the best of both worlds.
     - If False (default), will always use streaming case if available.
 
-    The main reason for this flag is that code might be written using ``.stream()`` and
+    The main reason for this flag is that code might be written using ``stream()`` and
     a user may want to swap out a given model for another model whose the implementation
     does not properly support streaming.
+
     """
 
     @model_validator(mode="before")
     @classmethod
     def raise_deprecation(cls, values: dict) -> Any:
-        """Raise deprecation warning if callback_manager is used.
+        """Raise deprecation warning if ``callback_manager`` is used.
 
         Args:
             values (Dict): Values to validate.
@@ -328,7 +332,8 @@ class BaseChatModel(BaseLanguageModel[BaseMessage], ABC):
             Dict: Validated values.
 
         Raises:
-            DeprecationWarning: If callback_manager is used.
+            DeprecationWarning: If ``callback_manager`` is used.
+
         """
         if values.get("callback_manager") is not None:
             warnings.warn(
@@ -653,6 +658,7 @@ class BaseChatModel(BaseLanguageModel[BaseMessage], ABC):
 
         Returns:
             List of ChatGeneration objects.
+
         """
         converted_generations = []
         for gen in cache_val:
@@ -778,7 +784,8 @@ class BaseChatModel(BaseLanguageModel[BaseMessage], ABC):
 
         Returns:
             An LLMResult, which contains a list of candidate Generations for each input
-                prompt and additional model provider-specific output.
+            prompt and additional model provider-specific output.
+
         """
         ls_structured_output_format = kwargs.pop(
             "ls_structured_output_format", None
@@ -892,7 +899,8 @@ class BaseChatModel(BaseLanguageModel[BaseMessage], ABC):
 
         Returns:
             An LLMResult, which contains a list of candidate Generations for each input
-                prompt and additional model provider-specific output.
+            prompt and additional model provider-specific output.
+
         """
         ls_structured_output_format = kwargs.pop(
             "ls_structured_output_format", None
@@ -1248,6 +1256,7 @@ class BaseChatModel(BaseLanguageModel[BaseMessage], ABC):
 
         Returns:
             The model output message.
+
         """
         generation = self.generate(
             [messages], stop=stop, callbacks=callbacks, **kwargs
@@ -1288,6 +1297,7 @@ class BaseChatModel(BaseLanguageModel[BaseMessage], ABC):
 
         Returns:
             The model output string.
+
         """
         return self.predict(message, stop=stop, **kwargs)
 
@@ -1307,6 +1317,7 @@ class BaseChatModel(BaseLanguageModel[BaseMessage], ABC):
 
         Returns:
             The predicted output string.
+
         """
         stop_ = None if stop is None else list(stop)
         result = self([HumanMessage(content=text)], stop=stop_, **kwargs)
@@ -1382,6 +1393,7 @@ class BaseChatModel(BaseLanguageModel[BaseMessage], ABC):
 
         Returns:
             A Runnable that returns a message.
+
         """
         raise NotImplementedError
 
@@ -1544,8 +1556,10 @@ class BaseChatModel(BaseLanguageModel[BaseMessage], ABC):
 class SimpleChatModel(BaseChatModel):
     """Simplified implementation for a chat model to inherit from.
 
-    **Note** This implementation is primarily here for backwards compatibility.
-        For new implementations, please use `BaseChatModel` directly.
+    .. note::
+        This implementation is primarily here for backwards compatibility. For new
+        implementations, please use ``BaseChatModel`` directly.
+
     """
 
     def _generate(

libs/core/langchain_core/language_models/fake_chat_models.py

@@ -3,7 +3,7 @@
 import asyncio
 import re
 import time
-from collections.abc import AsyncIterator, Iterator
+from collections.abc import AsyncIterator, Iterable, Iterator
 from typing import Any, Optional, Union, cast
 
 from typing_extensions import override
@@ -16,6 +16,10 @@ from langchain_core.language_models.chat_models import BaseChatModel, SimpleChat
 from langchain_core.messages import AIMessage, AIMessageChunk, BaseMessage
 from langchain_core.outputs import ChatGeneration, ChatGenerationChunk, ChatResult
 from langchain_core.runnables import RunnableConfig
+from langchain_core.v1.chat_models import BaseChatModel as BaseChatModelV1
+from langchain_core.v1.messages import AIMessage as AIMessageV1
+from langchain_core.v1.messages import AIMessageChunk as AIMessageChunkV1
+from langchain_core.v1.messages import MessageV1
 
 
 class FakeMessagesListChatModel(BaseChatModel):
@@ -223,11 +227,12 @@ class GenericFakeChatModel(BaseChatModel):
     This can be expanded to accept other types like Callables / dicts / strings
     to make the interface more generic if needed.
 
-    Note: if you want to pass a list, you can use `iter` to convert it to an iterator.
+    .. note::
+        if you want to pass a list, you can use ``iter`` to convert it to an iterator.
 
-    Please note that streaming is not implemented yet. We should try to implement it
-    in the future by delegating to invoke and then breaking the resulting output
-    into message chunks.
+    .. warning::
+        Streaming is not implemented yet. We should try to implement it in the future by
+        delegating to invoke and then breaking the resulting output into message chunks.
     """
 
     @override
@@ -367,3 +372,69 @@ class ParrotFakeChatModel(BaseChatModel):
     @property
     def _llm_type(self) -> str:
         return "parrot-fake-chat-model"
+
+
+class GenericFakeChatModelV1(BaseChatModelV1):
+    """Generic fake chat model that can be used to test the chat model interface."""
+
+    messages: Optional[Iterator[Union[AIMessageV1, str]]] = None
+    message_chunks: Optional[Iterable[Union[AIMessageChunkV1, str]]] = None
+
+    @override
+    def _invoke(
+        self,
+        messages: list[MessageV1],
+        run_manager: Optional[CallbackManagerForLLMRun] = None,
+        **kwargs: Any,
+    ) -> AIMessageV1:
+        """Top Level call."""
+        if self.messages is None:
+            error_msg = "Messages iterator is not set."
+            raise ValueError(error_msg)
+        message = next(self.messages)
+        return AIMessageV1(content=message) if isinstance(message, str) else message
+
+    @override
+    def _stream(
+        self,
+        messages: list[MessageV1],
+        run_manager: Optional[CallbackManagerForLLMRun] = None,
+        **kwargs: Any,
+    ) -> Iterator[AIMessageChunkV1]:
+        """Top Level call."""
+        if self.message_chunks is None:
+            error_msg = "Message chunks iterator is not set."
+            raise ValueError(error_msg)
+        for chunk in self.message_chunks:
+            if isinstance(chunk, str):
+                yield AIMessageChunkV1(chunk)
+            else:
+                yield chunk
+
+    @property
+    def _llm_type(self) -> str:
+        return "generic-fake-chat-model"
+
+
+class ParrotFakeChatModelV1(BaseChatModelV1):
+    """Generic fake chat model that can be used to test the chat model interface.
+
+    * Chat model should be usable in both sync and async tests
+    """
+
+    @override
+    def _invoke(
+        self,
+        messages: list[MessageV1],
+        stop: Optional[list[str]] = None,
+        run_manager: Optional[CallbackManagerForLLMRun] = None,
+        **kwargs: Any,
+    ) -> AIMessageV1:
+        """Top Level call."""
+        if isinstance(messages[-1], AIMessageV1):
+            return messages[-1]
+        return AIMessageV1(content=messages[-1].content)
+
+    @property
+    def _llm_type(self) -> str:
+        return "parrot-fake-chat-model"

libs/core/langchain_core/load/dump.py

@@ -1,11 +1,14 @@
 """Dump objects to json."""
 
+import dataclasses
+import inspect
 import json
 from typing import Any
 
 from pydantic import BaseModel
 
 from langchain_core.load.serializable import Serializable, to_json_not_implemented
+from langchain_core.v1.messages import MessageV1Types
 
 
 def default(obj: Any) -> Any:
@@ -19,6 +22,24 @@ def default(obj: Any) -> Any:
     """
     if isinstance(obj, Serializable):
         return obj.to_json()
+
+    # Handle v1 message classes
+    if type(obj) in MessageV1Types:
+        # Get the constructor signature to only include valid parameters
+        init_sig = inspect.signature(type(obj).__init__)
+        valid_params = set(init_sig.parameters.keys()) - {"self"}
+
+        # Filter dataclass fields to only include constructor params
+        all_fields = dataclasses.asdict(obj)
+        kwargs = {k: v for k, v in all_fields.items() if k in valid_params}
+
+        return {
+            "lc": 1,
+            "type": "constructor",
+            "id": ["langchain_core", "v1", "messages", type(obj).__name__],
+            "kwargs": kwargs,
+        }
+
     return to_json_not_implemented(obj)
 
 
@@ -73,10 +94,9 @@ def dumps(obj: Any, *, pretty: bool = False, **kwargs: Any) -> str:
 def dumpd(obj: Any) -> Any:
     """Return a dict representation of an object.
 
-    Note:
-        Unfortunately this function is not as efficient as it could be
-        because it first dumps the object to a json string and then loads it
-        back into a dictionary.
+    .. note::
+        Unfortunately this function is not as efficient as it could be because it first
+        dumps the object to a json string and then loads it back into a dictionary.
 
     Args:
         obj: The object to dump.

libs/core/langchain_core/load/load.py

@@ -156,8 +156,13 @@ class Reviver:
 
             cls = getattr(mod, name)
 
-            # The class must be a subclass of Serializable.
-            if not issubclass(cls, Serializable):
+            # Import MessageV1Types lazily to avoid circular import:
+            # load.load -> v1.messages -> messages.ai -> messages.base ->
+            #   load.serializable -> load.__init__ -> load.load
+            from langchain_core.v1.messages import MessageV1Types
+
+            # The class must be a subclass of Serializable or a v1 message class.
+            if not (issubclass(cls, Serializable) or cls in MessageV1Types):
                 msg = f"Invalid namespace: {value}"
                 raise ValueError(msg)
 

libs/core/langchain_core/messages/__init__.py

@@ -33,9 +33,31 @@ if TYPE_CHECKING:
     )
     from langchain_core.messages.chat import ChatMessage, ChatMessageChunk
     from langchain_core.messages.content_blocks import (
+        Annotation,
+        AudioContentBlock,
+        Citation,
+        CodeInterpreterCall,
+        CodeInterpreterOutput,
+        CodeInterpreterResult,
+        ContentBlock,
+        DataContentBlock,
+        FileContentBlock,
+        ImageContentBlock,
+        NonStandardAnnotation,
+        NonStandardContentBlock,
+        PlainTextContentBlock,
+        ReasoningContentBlock,
+        TextContentBlock,
+        VideoContentBlock,
+        WebSearchCall,
+        WebSearchResult,
         convert_to_openai_data_block,
         convert_to_openai_image_block,
         is_data_content_block,
+        is_reasoning_block,
+        is_text_block,
+        is_tool_call_block,
+        is_tool_call_chunk,
     )
     from langchain_core.messages.function import FunctionMessage, FunctionMessageChunk
     from langchain_core.messages.human import HumanMessage, HumanMessageChunk
@@ -65,24 +87,42 @@ if TYPE_CHECKING:
 __all__ = (
     "AIMessage",
     "AIMessageChunk",
+    "Annotation",
     "AnyMessage",
+    "AudioContentBlock",
     "BaseMessage",
     "BaseMessageChunk",
     "ChatMessage",
     "ChatMessageChunk",
+    "Citation",
+    "CodeInterpreterCall",
+    "CodeInterpreterOutput",
+    "CodeInterpreterResult",
+    "ContentBlock",
+    "DataContentBlock",
+    "FileContentBlock",
     "FunctionMessage",
     "FunctionMessageChunk",
     "HumanMessage",
     "HumanMessageChunk",
+    "ImageContentBlock",
     "InvalidToolCall",
     "MessageLikeRepresentation",
+    "NonStandardAnnotation",
+    "NonStandardContentBlock",
+    "PlainTextContentBlock",
+    "ReasoningContentBlock",
     "RemoveMessage",
     "SystemMessage",
     "SystemMessageChunk",
+    "TextContentBlock",
     "ToolCall",
     "ToolCallChunk",
     "ToolMessage",
     "ToolMessageChunk",
+    "VideoContentBlock",
+    "WebSearchCall",
+    "WebSearchResult",
     "_message_from_dict",
     "convert_to_messages",
     "convert_to_openai_data_block",
@@ -91,6 +131,10 @@ __all__ = (
     "filter_messages",
     "get_buffer_string",
     "is_data_content_block",
+    "is_reasoning_block",
+    "is_text_block",
+    "is_tool_call_block",
+    "is_tool_call_chunk",
     "merge_content",
     "merge_message_runs",
     "message_chunk_to_message",
@@ -103,25 +147,43 @@ __all__ = (
 _dynamic_imports = {
     "AIMessage": "ai",
     "AIMessageChunk": "ai",
+    "Annotation": "content_blocks",
+    "AudioContentBlock": "content_blocks",
     "BaseMessage": "base",
     "BaseMessageChunk": "base",
     "merge_content": "base",
     "message_to_dict": "base",
     "messages_to_dict": "base",
+    "Citation": "content_blocks",
+    "ContentBlock": "content_blocks",
     "ChatMessage": "chat",
     "ChatMessageChunk": "chat",
+    "CodeInterpreterCall": "content_blocks",
+    "CodeInterpreterOutput": "content_blocks",
+    "CodeInterpreterResult": "content_blocks",
+    "DataContentBlock": "content_blocks",
+    "FileContentBlock": "content_blocks",
     "FunctionMessage": "function",
     "FunctionMessageChunk": "function",
     "HumanMessage": "human",
     "HumanMessageChunk": "human",
+    "NonStandardAnnotation": "content_blocks",
+    "NonStandardContentBlock": "content_blocks",
+    "PlainTextContentBlock": "content_blocks",
+    "ReasoningContentBlock": "content_blocks",
     "RemoveMessage": "modifier",
     "SystemMessage": "system",
     "SystemMessageChunk": "system",
+    "WebSearchCall": "content_blocks",
+    "WebSearchResult": "content_blocks",
+    "ImageContentBlock": "content_blocks",
     "InvalidToolCall": "tool",
+    "TextContentBlock": "content_blocks",
     "ToolCall": "tool",
     "ToolCallChunk": "tool",
     "ToolMessage": "tool",
     "ToolMessageChunk": "tool",
+    "VideoContentBlock": "content_blocks",
     "AnyMessage": "utils",
     "MessageLikeRepresentation": "utils",
     "_message_from_dict": "utils",
@@ -132,6 +194,10 @@ _dynamic_imports = {
     "filter_messages": "utils",
     "get_buffer_string": "utils",
     "is_data_content_block": "content_blocks",
+    "is_reasoning_block": "content_blocks",
+    "is_text_block": "content_blocks",
+    "is_tool_call_block": "content_blocks",
+    "is_tool_call_chunk": "content_blocks",
     "merge_message_runs": "utils",
     "message_chunk_to_message": "utils",
     "messages_from_dict": "utils",

libs/core/langchain_core/messages/ai.py

@@ -8,11 +8,7 @@ from typing import Any, Literal, Optional, Union, cast
 from pydantic import model_validator
 from typing_extensions import NotRequired, Self, TypedDict, override
 
-from langchain_core.messages.base import (
-    BaseMessage,
-    BaseMessageChunk,
-    merge_content,
-)
+from langchain_core.messages.base import BaseMessage, BaseMessageChunk, merge_content
 from langchain_core.messages.tool import (
     InvalidToolCall,
     ToolCall,
@@ -20,23 +16,26 @@ from langchain_core.messages.tool import (
     default_tool_chunk_parser,
     default_tool_parser,
 )
-from langchain_core.messages.tool import (
-    invalid_tool_call as create_invalid_tool_call,
-)
-from langchain_core.messages.tool import (
-    tool_call as create_tool_call,
-)
-from langchain_core.messages.tool import (
-    tool_call_chunk as create_tool_call_chunk,
-)
+from langchain_core.messages.tool import invalid_tool_call as create_invalid_tool_call
+from langchain_core.messages.tool import tool_call as create_tool_call
+from langchain_core.messages.tool import tool_call_chunk as create_tool_call_chunk
 from langchain_core.utils._merge import merge_dicts, merge_lists
 from langchain_core.utils.json import parse_partial_json
 from langchain_core.utils.usage import _dict_int_op
 
 logger = logging.getLogger(__name__)
 
+_LC_AUTO_PREFIX = "lc_"
+"""LangChain auto-generated ID prefix for messages and content blocks."""
 
-_LC_ID_PREFIX = "run-"
+_LC_ID_PREFIX = f"{_LC_AUTO_PREFIX}run-"
+"""Internal tracing/callback system identifier.
+
+Used for:
+- Tracing. Every LangChain operation (LLM call, chain execution, tool use, etc.)
+  gets a unique run_id (UUID)
+- Enables tracking parent-child relationships between operations
+"""
 
 
 class InputTokenDetails(TypedDict, total=False):
@@ -428,17 +427,27 @@ def add_ai_message_chunks(
 
     chunk_id = None
     candidates = [left.id] + [o.id for o in others]
-    # first pass: pick the first non-run-* id
+    # first pass: pick the first provider-assigned id (non-run-* and non-lc_*)
     for id_ in candidates:
-        if id_ and not id_.startswith(_LC_ID_PREFIX):
+        if (
+            id_
+            and not id_.startswith(_LC_ID_PREFIX)
+            and not id_.startswith(_LC_AUTO_PREFIX)
+        ):
             chunk_id = id_
             break
     else:
-        # second pass: no provider-assigned id found, just take the first non-null
+        # second pass: prefer lc_run-* ids over lc_* ids
         for id_ in candidates:
-            if id_:
+            if id_ and id_.startswith(_LC_ID_PREFIX):
                 chunk_id = id_
                 break
+        else:
+            # third pass: take any remaining id (auto-generated lc_* ids)
+            for id_ in candidates:
+                if id_:
+                    chunk_id = id_
+                    break
 
     return left.__class__(
         example=left.example,

libs/core/langchain_core/messages/modifier.py

@@ -13,7 +13,7 @@ class RemoveMessage(BaseMessage):
 
     def __init__(
         self,
-        id: str,  # noqa: A002
+        id: str,
         **kwargs: Any,
     ) -> None:
         """Create a RemoveMessage.

libs/core/langchain_core/messages/tool.py

@@ -5,9 +5,12 @@ from typing import Any, Literal, Optional, Union
 from uuid import UUID
 
 from pydantic import Field, model_validator
-from typing_extensions import NotRequired, TypedDict, override
+from typing_extensions import override
 
 from langchain_core.messages.base import BaseMessage, BaseMessageChunk, merge_content
+from langchain_core.messages.content_blocks import InvalidToolCall as InvalidToolCall
+from langchain_core.messages.content_blocks import ToolCall as ToolCall
+from langchain_core.messages.content_blocks import ToolCallChunk as ToolCallChunk
 from langchain_core.utils._merge import merge_dicts, merge_obj
 
 
@@ -177,42 +180,11 @@ class ToolMessageChunk(ToolMessage, BaseMessageChunk):
         return super().__add__(other)
 
 
-class ToolCall(TypedDict):
-    """Represents a request to call a tool.
-
-    Example:
-
-        .. code-block:: python
-
-            {
-                "name": "foo",
-                "args": {"a": 1},
-                "id": "123"
-            }
-
-        This represents a request to call the tool named "foo" with arguments {"a": 1}
-        and an identifier of "123".
-
-    """
-
-    name: str
-    """The name of the tool to be called."""
-    args: dict[str, Any]
-    """The arguments to the tool call."""
-    id: Optional[str]
-    """An identifier associated with the tool call.
-
-    An identifier is needed to associate a tool call request with a tool
-    call result in events when multiple concurrent tool calls are made.
-    """
-    type: NotRequired[Literal["tool_call"]]
-
-
 def tool_call(
     *,
     name: str,
     args: dict[str, Any],
-    id: Optional[str],  # noqa: A002
+    id: Optional[str],
 ) -> ToolCall:
     """Create a tool call.
 
@@ -224,43 +196,11 @@ def tool_call(
     return ToolCall(name=name, args=args, id=id, type="tool_call")
 
 
-class ToolCallChunk(TypedDict):
-    """A chunk of a tool call (e.g., as part of a stream).
-
-    When merging ToolCallChunks (e.g., via AIMessageChunk.__add__),
-    all string attributes are concatenated. Chunks are only merged if their
-    values of `index` are equal and not None.
-
-    Example:
-
-    .. code-block:: python
-
-        left_chunks = [ToolCallChunk(name="foo", args='{"a":', index=0)]
-        right_chunks = [ToolCallChunk(name=None, args='1}', index=0)]
-
-        (
-            AIMessageChunk(content="", tool_call_chunks=left_chunks)
-            + AIMessageChunk(content="", tool_call_chunks=right_chunks)
-        ).tool_call_chunks == [ToolCallChunk(name='foo', args='{"a":1}', index=0)]
-
-    """
-
-    name: Optional[str]
-    """The name of the tool to be called."""
-    args: Optional[str]
-    """The arguments to the tool call."""
-    id: Optional[str]
-    """An identifier associated with the tool call."""
-    index: Optional[int]
-    """The index of the tool call in a sequence."""
-    type: NotRequired[Literal["tool_call_chunk"]]
-
-
 def tool_call_chunk(
     *,
     name: Optional[str] = None,
     args: Optional[str] = None,
-    id: Optional[str] = None,  # noqa: A002
+    id: Optional[str] = None,
     index: Optional[int] = None,
 ) -> ToolCallChunk:
     """Create a tool call chunk.
@@ -276,29 +216,11 @@ def tool_call_chunk(
     )
 
 
-class InvalidToolCall(TypedDict):
-    """Allowance for errors made by LLM.
-
-    Here we add an `error` key to surface errors made during generation
-    (e.g., invalid JSON arguments.)
-    """
-
-    name: Optional[str]
-    """The name of the tool to be called."""
-    args: Optional[str]
-    """The arguments to the tool call."""
-    id: Optional[str]
-    """An identifier associated with the tool call."""
-    error: Optional[str]
-    """An error message associated with the tool call."""
-    type: NotRequired[Literal["invalid_tool_call"]]
-
-
 def invalid_tool_call(
     *,
     name: Optional[str] = None,
     args: Optional[str] = None,
-    id: Optional[str] = None,  # noqa: A002
+    id: Optional[str] = None,
     error: Optional[str] = None,
 ) -> InvalidToolCall:
     """Create an invalid tool call.

libs/core/langchain_core/messages/utils.py

@@ -35,11 +35,18 @@ from langchain_core.messages import convert_to_openai_data_block, is_data_conten
 from langchain_core.messages.ai import AIMessage, AIMessageChunk
 from langchain_core.messages.base import BaseMessage, BaseMessageChunk
 from langchain_core.messages.chat import ChatMessage, ChatMessageChunk
+from langchain_core.messages.content_blocks import ContentBlock
 from langchain_core.messages.function import FunctionMessage, FunctionMessageChunk
 from langchain_core.messages.human import HumanMessage, HumanMessageChunk
 from langchain_core.messages.modifier import RemoveMessage
 from langchain_core.messages.system import SystemMessage, SystemMessageChunk
 from langchain_core.messages.tool import ToolCall, ToolMessage, ToolMessageChunk
+from langchain_core.v1.messages import AIMessage as AIMessageV1
+from langchain_core.v1.messages import AIMessageChunk as AIMessageChunkV1
+from langchain_core.v1.messages import HumanMessage as HumanMessageV1
+from langchain_core.v1.messages import MessageV1, MessageV1Types, ResponseMetadata
+from langchain_core.v1.messages import SystemMessage as SystemMessageV1
+from langchain_core.v1.messages import ToolMessage as ToolMessageV1
 
 if TYPE_CHECKING:
     from langchain_text_splitters import TextSplitter
@@ -203,7 +210,7 @@ def message_chunk_to_message(chunk: BaseMessageChunk) -> BaseMessage:
 
 
 MessageLikeRepresentation = Union[
-    BaseMessage, list[str], tuple[str, str], str, dict[str, Any]
+    BaseMessage, list[str], tuple[str, str], str, dict[str, Any], MessageV1
 ]
 
 
@@ -213,7 +220,7 @@ def _create_message_from_message_type(
     name: Optional[str] = None,
     tool_call_id: Optional[str] = None,
     tool_calls: Optional[list[dict[str, Any]]] = None,
-    id: Optional[str] = None,  # noqa: A002
+    id: Optional[str] = None,
     **additional_kwargs: Any,
 ) -> BaseMessage:
     """Create a message from a message type and content string.
@@ -294,6 +301,130 @@ def _create_message_from_message_type(
     return message
 
 
+def _create_message_from_message_type_v1(
+    message_type: str,
+    content: str,
+    name: Optional[str] = None,
+    tool_call_id: Optional[str] = None,
+    tool_calls: Optional[list[dict[str, Any]]] = None,
+    id: Optional[str] = None,
+    **kwargs: Any,
+) -> MessageV1:
+    """Create a message from a message type and content string.
+
+    Args:
+        message_type: (str) the type of the message (e.g., "human", "ai", etc.).
+        content: (str) the content string.
+        name: (str) the name of the message. Default is None.
+        tool_call_id: (str) the tool call id. Default is None.
+        tool_calls: (list[dict[str, Any]]) the tool calls. Default is None.
+        id: (str) the id of the message. Default is None.
+        kwargs: (dict[str, Any]) additional keyword arguments.
+
+    Returns:
+        a message of the appropriate type.
+
+    Raises:
+        ValueError: if the message type is not one of "human", "user", "ai",
+            "assistant", "tool", "system", or "developer".
+    """
+    if name is not None:
+        kwargs["name"] = name
+    if tool_call_id is not None:
+        kwargs["tool_call_id"] = tool_call_id
+    if kwargs and (response_metadata := kwargs.pop("response_metadata", None)):
+        kwargs["response_metadata"] = response_metadata
+    if id is not None:
+        kwargs["id"] = id
+    if tool_calls is not None:
+        kwargs["tool_calls"] = []
+        for tool_call in tool_calls:
+            # Convert OpenAI-format tool call to LangChain format.
+            if "function" in tool_call:
+                args = tool_call["function"]["arguments"]
+                if isinstance(args, str):
+                    args = json.loads(args, strict=False)
+                kwargs["tool_calls"].append(
+                    {
+                        "name": tool_call["function"]["name"],
+                        "args": args,
+                        "id": tool_call["id"],
+                        "type": "tool_call",
+                    }
+                )
+            else:
+                kwargs["tool_calls"].append(tool_call)
+    if message_type in {"human", "user"}:
+        message: MessageV1 = HumanMessageV1(content=content, **kwargs)
+    elif message_type in {"ai", "assistant"}:
+        message = AIMessageV1(content=content, **kwargs)
+    elif message_type in {"system", "developer"}:
+        if message_type == "developer":
+            kwargs["custom_role"] = "developer"
+        message = SystemMessageV1(content=content, **kwargs)
+    elif message_type == "tool":
+        artifact = kwargs.pop("artifact", None)
+        message = ToolMessageV1(content=content, artifact=artifact, **kwargs)
+    else:
+        msg = (
+            f"Unexpected message type: '{message_type}'. Use one of 'human',"
+            f" 'user', 'ai', 'assistant', 'function', 'tool', 'system', or 'developer'."
+        )
+        msg = create_message(message=msg, error_code=ErrorCode.MESSAGE_COERCION_FAILURE)
+        raise ValueError(msg)
+    return message
+
+
+def convert_from_v1_message(message: MessageV1) -> BaseMessage:
+    """Compatibility layer to convert v1 messages to current messages.
+
+    Args:
+        message: MessageV1 instance to convert.
+
+    Returns:
+        BaseMessage: Converted message instance.
+    """
+    content = cast("Union[str, list[str | dict]]", message.content)
+    if isinstance(message, AIMessageV1):
+        return AIMessage(
+            content=content,
+            id=message.id,
+            name=message.name,
+            tool_calls=message.tool_calls,
+            response_metadata=cast("dict", message.response_metadata),
+        )
+    if isinstance(message, AIMessageChunkV1):
+        return AIMessageChunk(
+            content=content,
+            id=message.id,
+            name=message.name,
+            tool_call_chunks=message.tool_call_chunks,
+            response_metadata=cast("dict", message.response_metadata),
+        )
+    if isinstance(message, HumanMessageV1):
+        return HumanMessage(
+            content=content,
+            id=message.id,
+            name=message.name,
+        )
+    if isinstance(message, SystemMessageV1):
+        return SystemMessage(
+            content=content,
+            id=message.id,
+        )
+    if isinstance(message, ToolMessageV1):
+        return ToolMessage(
+            content=content,
+            id=message.id,
+            tool_call_id=message.tool_call_id,
+            artifact=message.artifact,
+            name=message.name,
+            status=message.status,
+        )
+    message = f"Unsupported message type: {type(message)}"
+    raise NotImplementedError(message)
+
+
 def _convert_to_message(message: MessageLikeRepresentation) -> BaseMessage:
     """Instantiate a message from a variety of message formats.
 
@@ -341,6 +472,143 @@ def _convert_to_message(message: MessageLikeRepresentation) -> BaseMessage:
         message_ = _create_message_from_message_type(
             msg_type, msg_content, **msg_kwargs
         )
+    elif isinstance(message, MessageV1Types):
+        message_ = convert_from_v1_message(message)
+    else:
+        msg = f"Unsupported message type: {type(message)}"
+        msg = create_message(message=msg, error_code=ErrorCode.MESSAGE_COERCION_FAILURE)
+        raise NotImplementedError(msg)
+
+    return message_
+
+
+def _convert_from_v0_to_v1(message: BaseMessage) -> MessageV1:
+    """Convert a v0 message to a v1 message."""
+    if isinstance(message, HumanMessage):  # Checking for v0 HumanMessage
+        return HumanMessageV1(message.content, id=message.id, name=message.name)  # type: ignore[arg-type]
+    if isinstance(message, AIMessage):  # Checking for v0 AIMessage
+        return AIMessageV1(
+            content=message.content,  # type: ignore[arg-type]
+            id=message.id,
+            name=message.name,
+            lc_version="v1",
+            response_metadata=message.response_metadata,  # type: ignore[arg-type]
+            usage_metadata=message.usage_metadata,
+            tool_calls=message.tool_calls,
+            invalid_tool_calls=message.invalid_tool_calls,
+        )
+    if isinstance(message, SystemMessage):  # Checking for v0 SystemMessage
+        return SystemMessageV1(
+            message.content,  # type: ignore[arg-type]
+            id=message.id,
+            name=message.name,
+        )
+    if isinstance(message, ToolMessage):  # Checking for v0 ToolMessage
+        return ToolMessageV1(
+            message.content,  # type: ignore[arg-type]
+            message.tool_call_id,
+            id=message.id,
+            name=message.name,
+            artifact=message.artifact,
+            status=message.status,
+        )
+    msg = f"Unsupported v0 message type for conversion to v1: {type(message)}"
+    raise NotImplementedError(msg)
+
+
+def _safe_convert_from_v0_to_v1(message: BaseMessage) -> MessageV1:
+    """Convert a v0 message to a v1 message."""
+    from langchain_core.messages.content_blocks import create_text_block
+
+    if isinstance(message, HumanMessage):  # Checking for v0 HumanMessage
+        content: list[ContentBlock] = [create_text_block(str(message.content))]
+        return HumanMessageV1(content, id=message.id, name=message.name)
+    if isinstance(message, AIMessage):  # Checking for v0 AIMessage
+        content = [create_text_block(str(message.content))]
+
+        # Construct ResponseMetadata TypedDict from v0 response_metadata dict
+        # Since ResponseMetadata has total=False, we can safely cast the dict
+        response_metadata = cast("ResponseMetadata", message.response_metadata or {})
+        return AIMessageV1(
+            content=content,
+            id=message.id,
+            name=message.name,
+            lc_version="v1",
+            response_metadata=response_metadata,
+            usage_metadata=message.usage_metadata,
+            tool_calls=message.tool_calls,
+            invalid_tool_calls=message.invalid_tool_calls,
+        )
+    if isinstance(message, SystemMessage):  # Checking for v0 SystemMessage
+        content = [create_text_block(str(message.content))]
+        return SystemMessageV1(content=content, id=message.id, name=message.name)
+    if isinstance(message, ToolMessage):  # Checking for v0 ToolMessage
+        content = [create_text_block(str(message.content))]
+        return ToolMessageV1(
+            content,
+            message.tool_call_id,
+            id=message.id,
+            name=message.name,
+            artifact=message.artifact,
+            status=message.status,
+        )
+    msg = f"Unsupported v0 message type for conversion to v1: {type(message)}"
+    raise NotImplementedError(msg)
+
+
+def _convert_to_message_v1(message: MessageLikeRepresentation) -> MessageV1:
+    """Instantiate a message from a variety of message formats.
+
+    The message format can be one of the following:
+
+    - BaseMessagePromptTemplate
+    - BaseMessage
+    - 2-tuple of (role string, template); e.g., ("human", "{user_input}")
+    - dict: a message dict with role and content keys
+    - string: shorthand for ("human", template); e.g., "{user_input}"
+
+    Args:
+        message: a representation of a message in one of the supported formats.
+
+    Returns:
+        an instance of a message or a message template.
+
+    Raises:
+        NotImplementedError: if the message type is not supported.
+        ValueError: if the message dict does not contain the required keys.
+    """
+    if isinstance(message, MessageV1Types):
+        if isinstance(message, AIMessageChunkV1):
+            message_: MessageV1 = message.to_message()
+        else:
+            message_ = message
+    elif isinstance(message, BaseMessage):
+        # Convert v0 messages to v1 messages
+        message_ = _convert_from_v0_to_v1(message)
+    elif isinstance(message, str):
+        message_ = _create_message_from_message_type_v1("human", message)
+    elif isinstance(message, Sequence) and len(message) == 2:
+        # mypy doesn't realise this can't be a string given the previous branch
+        message_type_str, template = message  # type: ignore[misc]
+        message_ = _create_message_from_message_type_v1(message_type_str, template)
+    elif isinstance(message, dict):
+        msg_kwargs = message.copy()
+        try:
+            try:
+                msg_type = msg_kwargs.pop("role")
+            except KeyError:
+                msg_type = msg_kwargs.pop("type")
+            # None msg content is not allowed
+            msg_content = msg_kwargs.pop("content") or ""
+        except KeyError as e:
+            msg = f"Message dict must contain 'role' and 'content' keys, got {message}"
+            msg = create_message(
+                message=msg, error_code=ErrorCode.MESSAGE_COERCION_FAILURE
+            )
+            raise ValueError(msg) from e
+        message_ = _create_message_from_message_type_v1(
+            msg_type, msg_content, **msg_kwargs
+        )
     else:
         msg = f"Unsupported message type: {type(message)}"
         msg = create_message(message=msg, error_code=ErrorCode.MESSAGE_COERCION_FAILURE)
@@ -368,6 +636,25 @@ def convert_to_messages(
     return [_convert_to_message(m) for m in messages]
 
 
+def convert_to_messages_v1(
+    messages: Union[Iterable[MessageLikeRepresentation], PromptValue],
+) -> list[MessageV1]:
+    """Convert a sequence of messages to a list of messages.
+
+    Args:
+        messages: Sequence of messages to convert.
+
+    Returns:
+        list of messages (BaseMessages).
+    """
+    # Import here to avoid circular imports
+    from langchain_core.prompt_values import PromptValue
+
+    if isinstance(messages, PromptValue):
+        return messages.to_messages(message_version="v1")
+    return [_convert_to_message_v1(m) for m in messages]
+
+
 def _runnable_support(func: Callable) -> Callable:
     @overload
     def wrapped(
@@ -656,22 +943,23 @@ def trim_messages(
     properties:
 
     1. The resulting chat history should be valid. Most chat models expect that chat
-       history starts with either (1) a `HumanMessage` or (2) a `SystemMessage` followed
-       by a `HumanMessage`. To achieve this, set `start_on="human"`.
-       In addition, generally a `ToolMessage` can only appear after an `AIMessage`
+       history starts with either (1) a ``HumanMessage`` or (2) a ``SystemMessage`` followed
+       by a ``HumanMessage``. To achieve this, set ``start_on="human"``.
+       In addition, generally a ``ToolMessage`` can only appear after an ``AIMessage``
        that involved a tool call.
        Please see the following link for more information about messages:
        https://python.langchain.com/docs/concepts/#messages
     2. It includes recent messages and drops old messages in the chat history.
-       To achieve this set the `strategy="last"`.
-    3. Usually, the new chat history should include the `SystemMessage` if it
-       was present in the original chat history since the `SystemMessage` includes
-       special instructions to the chat model. The `SystemMessage` is almost always
+       To achieve this set the ``strategy="last"``.
+    3. Usually, the new chat history should include the ``SystemMessage`` if it
+       was present in the original chat history since the ``SystemMessage`` includes
+       special instructions to the chat model. The ``SystemMessage`` is almost always
        the first message in the history if present. To achieve this set the
-       `include_system=True`.
+       ``include_system=True``.
 
-    **Note** The examples below show how to configure `trim_messages` to achieve
-        a behavior consistent with the above properties.
+    .. note::
+        The examples below show how to configure ``trim_messages`` to achieve a behavior
+        consistent with the above properties.
 
     Args:
         messages: Sequence of Message-like objects to trim.
@@ -1007,10 +1295,11 @@ def convert_to_openai_messages(
 
     oai_messages: list = []
 
-    if is_single := isinstance(messages, (BaseMessage, dict, str)):
+    if is_single := isinstance(messages, (BaseMessage, dict, str, MessageV1Types)):
         messages = [messages]
 
-    messages = convert_to_messages(messages)
+    # TODO: resolve type ignore here
+    messages = convert_to_messages(messages)  # type: ignore[arg-type]
 
     for i, message in enumerate(messages):
         oai_msg: dict = {"role": _get_message_openai_role(message)}
@@ -1580,26 +1869,26 @@ def count_tokens_approximately(
         chars_per_token: Number of characters per token to use for the approximation.
             Default is 4 (one token corresponds to ~4 chars for common English text).
             You can also specify float values for more fine-grained control.
-            See more here: https://platform.openai.com/tokenizer
+            `See more here. <https://platform.openai.com/tokenizer>`__
         extra_tokens_per_message: Number of extra tokens to add per message.
             Default is 3 (special tokens, including beginning/end of message).
             You can also specify float values for more fine-grained control.
-            See more here:
-            https://github.com/openai/openai-cookbook/blob/main/examples/How_to_count_tokens_with_tiktoken.ipynb
+            `See more here. <https://github.com/openai/openai-cookbook/blob/main/examples/How_to_count_tokens_with_tiktoken.ipynb>`__
         count_name: Whether to include message names in the count.
             Enabled by default.
 
     Returns:
         Approximate number of tokens in the messages.
 
-    Note:
-        This is a simple approximation that may not match the exact token count
-        used by specific models. For accurate counts, use model-specific tokenizers.
+    .. note::
+        This is a simple approximation that may not match the exact token count used by
+        specific models. For accurate counts, use model-specific tokenizers.
 
     Warning:
         This function does not currently support counting image tokens.
 
     .. versionadded:: 0.3.46
+
     """
     token_count = 0.0
     for message in convert_to_messages(messages):

libs/core/langchain_core/output_parsers/base.py

@@ -11,6 +11,7 @@ from typing import (
     Optional,
     TypeVar,
     Union,
+    cast,
 )
 
 from typing_extensions import override
@@ -20,19 +21,22 @@ from langchain_core.messages import AnyMessage, BaseMessage
 from langchain_core.outputs import ChatGeneration, Generation
 from langchain_core.runnables import Runnable, RunnableConfig, RunnableSerializable
 from langchain_core.runnables.config import run_in_executor
+from langchain_core.v1.messages import AIMessage, MessageV1, MessageV1Types
 
 if TYPE_CHECKING:
     from langchain_core.prompt_values import PromptValue
 
 T = TypeVar("T")
-OutputParserLike = Runnable[LanguageModelOutput, T]
+OutputParserLike = Runnable[Union[LanguageModelOutput, AIMessage], T]
 
 
 class BaseLLMOutputParser(ABC, Generic[T]):
     """Abstract base class for parsing the outputs of a model."""
 
     @abstractmethod
-    def parse_result(self, result: list[Generation], *, partial: bool = False) -> T:
+    def parse_result(
+        self, result: Union[list[Generation], AIMessage], *, partial: bool = False
+    ) -> T:
         """Parse a list of candidate model Generations into a specific format.
 
         Args:
@@ -46,7 +50,7 @@ class BaseLLMOutputParser(ABC, Generic[T]):
         """
 
     async def aparse_result(
-        self, result: list[Generation], *, partial: bool = False
+        self, result: Union[list[Generation], AIMessage], *, partial: bool = False
     ) -> T:
         """Async parse a list of candidate model Generations into a specific format.
 
@@ -71,7 +75,7 @@ class BaseGenerationOutputParser(
     @override
     def InputType(self) -> Any:
         """Return the input type for the parser."""
-        return Union[str, AnyMessage]
+        return Union[str, AnyMessage, MessageV1]
 
     @property
     @override
@@ -84,7 +88,7 @@ class BaseGenerationOutputParser(
     @override
     def invoke(
         self,
-        input: Union[str, BaseMessage],
+        input: Union[str, BaseMessage, MessageV1],
         config: Optional[RunnableConfig] = None,
         **kwargs: Any,
     ) -> T:
@@ -97,9 +101,16 @@ class BaseGenerationOutputParser(
                 config,
                 run_type="parser",
             )
+        if isinstance(input, MessageV1Types):
+            return self._call_with_config(
+                lambda inner_input: self.parse_result(inner_input),
+                input,
+                config,
+                run_type="parser",
+            )
         return self._call_with_config(
             lambda inner_input: self.parse_result([Generation(text=inner_input)]),
-            input,
+            cast("str", input),
             config,
             run_type="parser",
         )
@@ -120,6 +131,13 @@ class BaseGenerationOutputParser(
                 config,
                 run_type="parser",
             )
+        if isinstance(input, MessageV1Types):
+            return await self._acall_with_config(
+                lambda inner_input: self.aparse_result(inner_input),
+                input,
+                config,
+                run_type="parser",
+            )
         return await self._acall_with_config(
             lambda inner_input: self.aparse_result([Generation(text=inner_input)]),
             input,
@@ -129,7 +147,7 @@ class BaseGenerationOutputParser(
 
 
 class BaseOutputParser(
-    BaseLLMOutputParser, RunnableSerializable[LanguageModelOutput, T]
+    BaseLLMOutputParser, RunnableSerializable[Union[LanguageModelOutput, AIMessage], T]
 ):
     """Base class to parse the output of an LLM call.
 
@@ -162,7 +180,7 @@ class BaseOutputParser(
     @override
     def InputType(self) -> Any:
         """Return the input type for the parser."""
-        return Union[str, AnyMessage]
+        return Union[str, AnyMessage, MessageV1]
 
     @property
     @override
@@ -189,7 +207,7 @@ class BaseOutputParser(
     @override
     def invoke(
         self,
-        input: Union[str, BaseMessage],
+        input: Union[str, BaseMessage, MessageV1],
         config: Optional[RunnableConfig] = None,
         **kwargs: Any,
     ) -> T:
@@ -202,9 +220,16 @@ class BaseOutputParser(
                 config,
                 run_type="parser",
             )
+        if isinstance(input, MessageV1Types):
+            return self._call_with_config(
+                lambda inner_input: self.parse_result(inner_input),
+                input,
+                config,
+                run_type="parser",
+            )
         return self._call_with_config(
             lambda inner_input: self.parse_result([Generation(text=inner_input)]),
-            input,
+            cast("str", input),
             config,
             run_type="parser",
         )
@@ -212,7 +237,7 @@ class BaseOutputParser(
     @override
     async def ainvoke(
         self,
-        input: Union[str, BaseMessage],
+        input: Union[str, BaseMessage, MessageV1],
         config: Optional[RunnableConfig] = None,
         **kwargs: Optional[Any],
     ) -> T:
@@ -225,15 +250,24 @@ class BaseOutputParser(
                 config,
                 run_type="parser",
             )
+        if isinstance(input, MessageV1Types):
+            return await self._acall_with_config(
+                lambda inner_input: self.aparse_result(inner_input),
+                input,
+                config,
+                run_type="parser",
+            )
         return await self._acall_with_config(
             lambda inner_input: self.aparse_result([Generation(text=inner_input)]),
-            input,
+            cast("str", input),
             config,
             run_type="parser",
         )
 
     @override
-    def parse_result(self, result: list[Generation], *, partial: bool = False) -> T:
+    def parse_result(
+        self, result: Union[list[Generation], AIMessage], *, partial: bool = False
+    ) -> T:
         """Parse a list of candidate model Generations into a specific format.
 
         The return value is parsed from only the first Generation in the result, which
@@ -248,6 +282,8 @@ class BaseOutputParser(
         Returns:
             Structured output.
         """
+        if isinstance(result, AIMessage):
+            return self.parse(result.text)
         return self.parse(result[0].text)
 
     @abstractmethod
@@ -262,7 +298,7 @@ class BaseOutputParser(
         """
 
     async def aparse_result(
-        self, result: list[Generation], *, partial: bool = False
+        self, result: Union[list[Generation], AIMessage], *, partial: bool = False
     ) -> T:
         """Async parse a list of candidate model Generations into a specific format.
 

libs/core/langchain_core/output_parsers/json.py

@@ -21,6 +21,7 @@ from langchain_core.utils.json import (
     parse_json_markdown,
     parse_partial_json,
 )
+from langchain_core.v1.messages import AIMessage
 
 # Union type needs to be last assignment to PydanticBaseModel to make mypy happy.
 PydanticBaseModel = Union[BaseModel, pydantic.BaseModel]
@@ -53,7 +54,9 @@ class JsonOutputParser(BaseCumulativeTransformOutputParser[Any]):
             return pydantic_object.schema()
         return None
 
-    def parse_result(self, result: list[Generation], *, partial: bool = False) -> Any:
+    def parse_result(
+        self, result: Union[list[Generation], AIMessage], *, partial: bool = False
+    ) -> Any:
         """Parse the result of an LLM call to a JSON object.
 
         Args:
@@ -70,7 +73,7 @@ class JsonOutputParser(BaseCumulativeTransformOutputParser[Any]):
         Raises:
             OutputParserException: If the output is not valid JSON.
         """
-        text = result[0].text
+        text = result.text if isinstance(result, AIMessage) else result[0].text
         text = text.strip()
         if partial:
             try:

libs/core/langchain_core/output_parsers/list.py

@@ -13,6 +13,7 @@ from typing_extensions import override
 
 from langchain_core.messages import BaseMessage
 from langchain_core.output_parsers.transform import BaseTransformOutputParser
+from langchain_core.v1.messages import AIMessage
 
 if TYPE_CHECKING:
     from collections.abc import AsyncIterator, Iterator
@@ -71,7 +72,7 @@ class ListOutputParser(BaseTransformOutputParser[list[str]]):
 
     @override
     def _transform(
-        self, input: Iterator[Union[str, BaseMessage]]
+        self, input: Iterator[Union[str, BaseMessage, AIMessage]]
     ) -> Iterator[list[str]]:
         buffer = ""
         for chunk in input:
@@ -81,6 +82,8 @@ class ListOutputParser(BaseTransformOutputParser[list[str]]):
                 if not isinstance(chunk_content, str):
                     continue
                 buffer += chunk_content
+            elif isinstance(chunk, AIMessage):
+                buffer += chunk.text
             else:
                 # add current chunk to buffer
                 buffer += chunk
@@ -105,7 +108,7 @@ class ListOutputParser(BaseTransformOutputParser[list[str]]):
 
     @override
     async def _atransform(
-        self, input: AsyncIterator[Union[str, BaseMessage]]
+        self, input: AsyncIterator[Union[str, BaseMessage, AIMessage]]
     ) -> AsyncIterator[list[str]]:
         buffer = ""
         async for chunk in input:
@@ -115,6 +118,8 @@ class ListOutputParser(BaseTransformOutputParser[list[str]]):
                 if not isinstance(chunk_content, str):
                     continue
                 buffer += chunk_content
+            elif isinstance(chunk, AIMessage):
+                buffer += chunk.text
             else:
                 # add current chunk to buffer
                 buffer += chunk

libs/core/langchain_core/output_parsers/openai_functions.py

@@ -17,6 +17,7 @@ from langchain_core.output_parsers import (
 )
 from langchain_core.output_parsers.json import parse_partial_json
 from langchain_core.outputs import ChatGeneration, Generation
+from langchain_core.v1.messages import AIMessage
 
 
 class OutputFunctionsParser(BaseGenerationOutputParser[Any]):
@@ -26,7 +27,9 @@ class OutputFunctionsParser(BaseGenerationOutputParser[Any]):
     """Whether to only return the arguments to the function call."""
 
     @override
-    def parse_result(self, result: list[Generation], *, partial: bool = False) -> Any:
+    def parse_result(
+        self, result: Union[list[Generation], AIMessage], *, partial: bool = False
+    ) -> Any:
         """Parse the result of an LLM call to a JSON object.
 
         Args:
@@ -39,6 +42,12 @@ class OutputFunctionsParser(BaseGenerationOutputParser[Any]):
         Raises:
             OutputParserException: If the output is not valid JSON.
         """
+        if isinstance(result, AIMessage):
+            msg = (
+                "This output parser does not support v1 AIMessages. Use "
+                "JsonOutputToolsParser instead."
+            )
+            raise TypeError(msg)
         generation = result[0]
         if not isinstance(generation, ChatGeneration):
             msg = "This output parser can only be used with a chat generation."
@@ -77,7 +86,9 @@ class JsonOutputFunctionsParser(BaseCumulativeTransformOutputParser[Any]):
     def _diff(self, prev: Optional[Any], next: Any) -> Any:
         return jsonpatch.make_patch(prev, next).patch
 
-    def parse_result(self, result: list[Generation], *, partial: bool = False) -> Any:
+    def parse_result(
+        self, result: Union[list[Generation], AIMessage], *, partial: bool = False
+    ) -> Any:
         """Parse the result of an LLM call to a JSON object.
 
         Args:
@@ -90,6 +101,12 @@ class JsonOutputFunctionsParser(BaseCumulativeTransformOutputParser[Any]):
         Raises:
             OutputParserException: If the output is not valid JSON.
         """
+        if isinstance(result, AIMessage):
+            msg = (
+                "This output parser does not support v1 AIMessages. Use "
+                "JsonOutputToolsParser instead."
+            )
+            raise TypeError(msg)
         if len(result) != 1:
             msg = f"Expected exactly one result, but got {len(result)}"
             raise OutputParserException(msg)
@@ -160,7 +177,9 @@ class JsonKeyOutputFunctionsParser(JsonOutputFunctionsParser):
     key_name: str
     """The name of the key to return."""
 
-    def parse_result(self, result: list[Generation], *, partial: bool = False) -> Any:
+    def parse_result(
+        self, result: Union[list[Generation], AIMessage], *, partial: bool = False
+    ) -> Any:
         """Parse the result of an LLM call to a JSON object.
 
         Args:
@@ -254,7 +273,9 @@ class PydanticOutputFunctionsParser(OutputFunctionsParser):
         return values
 
     @override
-    def parse_result(self, result: list[Generation], *, partial: bool = False) -> Any:
+    def parse_result(
+        self, result: Union[list[Generation], AIMessage], *, partial: bool = False
+    ) -> Any:
         """Parse the result of an LLM call to a JSON object.
 
         Args:
@@ -294,7 +315,9 @@ class PydanticAttrOutputFunctionsParser(PydanticOutputFunctionsParser):
     """The name of the attribute to return."""
 
     @override
-    def parse_result(self, result: list[Generation], *, partial: bool = False) -> Any:
+    def parse_result(
+        self, result: Union[list[Generation], AIMessage], *, partial: bool = False
+    ) -> Any:
         """Parse the result of an LLM call to a JSON object.
 
         Args:

libs/core/langchain_core/output_parsers/openai_tools.py

@@ -4,7 +4,7 @@ import copy
 import json
 import logging
 from json import JSONDecodeError
-from typing import Annotated, Any, Optional
+from typing import Annotated, Any, Optional, Union
 
 from pydantic import SkipValidation, ValidationError
 
@@ -16,6 +16,7 @@ from langchain_core.output_parsers.transform import BaseCumulativeTransformOutpu
 from langchain_core.outputs import ChatGeneration, Generation
 from langchain_core.utils.json import parse_partial_json
 from langchain_core.utils.pydantic import TypeBaseModel
+from langchain_core.v1.messages import AIMessage as AIMessageV1
 
 logger = logging.getLogger(__name__)
 
@@ -156,7 +157,9 @@ class JsonOutputToolsParser(BaseCumulativeTransformOutputParser[Any]):
     If no tool calls are found, None will be returned.
     """
 
-    def parse_result(self, result: list[Generation], *, partial: bool = False) -> Any:
+    def parse_result(
+        self, result: Union[list[Generation], AIMessageV1], *, partial: bool = False
+    ) -> Any:
         """Parse the result of an LLM call to a list of tool calls.
 
         Args:
@@ -173,31 +176,45 @@ class JsonOutputToolsParser(BaseCumulativeTransformOutputParser[Any]):
         Raises:
             OutputParserException: If the output is not valid JSON.
         """
-        generation = result[0]
-        if not isinstance(generation, ChatGeneration):
-            msg = "This output parser can only be used with a chat generation."
-            raise OutputParserException(msg)
-        message = generation.message
-        if isinstance(message, AIMessage) and message.tool_calls:
-            tool_calls = [dict(tc) for tc in message.tool_calls]
+        if isinstance(result, list):
+            generation = result[0]
+            if not isinstance(generation, ChatGeneration):
+                msg = (
+                    "This output parser can only be used with a chat generation or "
+                    "v1 AIMessage."
+                )
+                raise OutputParserException(msg)
+            message = generation.message
+            if isinstance(message, AIMessage) and message.tool_calls:
+                tool_calls = [dict(tc) for tc in message.tool_calls]
+                for tool_call in tool_calls:
+                    if not self.return_id:
+                        _ = tool_call.pop("id")
+            else:
+                try:
+                    raw_tool_calls = copy.deepcopy(
+                        message.additional_kwargs["tool_calls"]
+                    )
+                except KeyError:
+                    return []
+                tool_calls = parse_tool_calls(
+                    raw_tool_calls,
+                    partial=partial,
+                    strict=self.strict,
+                    return_id=self.return_id,
+                )
+        elif result.tool_calls:
+            # v1 message
+            tool_calls = [dict(tc) for tc in result.tool_calls]
             for tool_call in tool_calls:
                 if not self.return_id:
                     _ = tool_call.pop("id")
         else:
-            try:
-                raw_tool_calls = copy.deepcopy(message.additional_kwargs["tool_calls"])
-            except KeyError:
-                return []
-            tool_calls = parse_tool_calls(
-                raw_tool_calls,
-                partial=partial,
-                strict=self.strict,
-                return_id=self.return_id,
-            )
+            return []
+
         # for backwards compatibility
         for tc in tool_calls:
             tc["type"] = tc.pop("name")
-
         if self.first_tool_only:
             return tool_calls[0] if tool_calls else None
         return tool_calls
@@ -220,7 +237,9 @@ class JsonOutputKeyToolsParser(JsonOutputToolsParser):
     key_name: str
     """The type of tools to return."""
 
-    def parse_result(self, result: list[Generation], *, partial: bool = False) -> Any:
+    def parse_result(
+        self, result: Union[list[Generation], AIMessageV1], *, partial: bool = False
+    ) -> Any:
         """Parse the result of an LLM call to a list of tool calls.
 
         Args:
@@ -234,32 +253,47 @@ class JsonOutputKeyToolsParser(JsonOutputToolsParser):
         Returns:
             The parsed tool calls.
         """
-        generation = result[0]
-        if not isinstance(generation, ChatGeneration):
-            msg = "This output parser can only be used with a chat generation."
-            raise OutputParserException(msg)
-        message = generation.message
-        if isinstance(message, AIMessage) and message.tool_calls:
-            parsed_tool_calls = [dict(tc) for tc in message.tool_calls]
+        if isinstance(result, list):
+            generation = result[0]
+            if not isinstance(generation, ChatGeneration):
+                msg = "This output parser can only be used with a chat generation."
+                raise OutputParserException(msg)
+            message = generation.message
+            if isinstance(message, AIMessage) and message.tool_calls:
+                parsed_tool_calls = [dict(tc) for tc in message.tool_calls]
+                for tool_call in parsed_tool_calls:
+                    if not self.return_id:
+                        _ = tool_call.pop("id")
+            else:
+                try:
+                    raw_tool_calls = copy.deepcopy(
+                        message.additional_kwargs["tool_calls"]
+                    )
+                except KeyError:
+                    if self.first_tool_only:
+                        return None
+                    return []
+                parsed_tool_calls = parse_tool_calls(
+                    raw_tool_calls,
+                    partial=partial,
+                    strict=self.strict,
+                    return_id=self.return_id,
+                )
+        elif result.tool_calls:
+            # v1 message
+            parsed_tool_calls = [dict(tc) for tc in result.tool_calls]
             for tool_call in parsed_tool_calls:
                 if not self.return_id:
                     _ = tool_call.pop("id")
         else:
-            try:
-                raw_tool_calls = copy.deepcopy(message.additional_kwargs["tool_calls"])
-            except KeyError:
-                if self.first_tool_only:
-                    return None
-                return []
-            parsed_tool_calls = parse_tool_calls(
-                raw_tool_calls,
-                partial=partial,
-                strict=self.strict,
-                return_id=self.return_id,
-            )
+            if self.first_tool_only:
+                return None
+            return []
+
         # For backwards compatibility
         for tc in parsed_tool_calls:
             tc["type"] = tc.pop("name")
+
         if self.first_tool_only:
             parsed_result = list(
                 filter(lambda x: x["type"] == self.key_name, parsed_tool_calls)
@@ -299,7 +333,9 @@ class PydanticToolsParser(JsonOutputToolsParser):
 
     # TODO: Support more granular streaming of objects. Currently only streams once all
     # Pydantic object fields are present.
-    def parse_result(self, result: list[Generation], *, partial: bool = False) -> Any:
+    def parse_result(
+        self, result: Union[list[Generation], AIMessageV1], *, partial: bool = False
+    ) -> Any:
         """Parse the result of an LLM call to a list of Pydantic objects.
 
         Args:
@@ -337,12 +373,19 @@ class PydanticToolsParser(JsonOutputToolsParser):
             except (ValidationError, ValueError):
                 if partial:
                     continue
-                has_max_tokens_stop_reason = any(
-                    generation.message.response_metadata.get("stop_reason")
-                    == "max_tokens"
-                    for generation in result
-                    if isinstance(generation, ChatGeneration)
-                )
+                has_max_tokens_stop_reason = False
+                if isinstance(result, list):
+                    has_max_tokens_stop_reason = any(
+                        generation.message.response_metadata.get("stop_reason")
+                        == "max_tokens"
+                        for generation in result
+                        if isinstance(generation, ChatGeneration)
+                    )
+                else:
+                    # v1 message
+                    has_max_tokens_stop_reason = (
+                        result.response_metadata.get("stop_reason") == "max_tokens"
+                    )
                 if has_max_tokens_stop_reason:
                     logger.exception(_MAX_TOKENS_ERROR)
                 raise

libs/core/langchain_core/output_parsers/pydantic.py

@@ -1,7 +1,7 @@
 """Output parsers using Pydantic."""
 
 import json
-from typing import Annotated, Generic, Optional
+from typing import Annotated, Generic, Optional, Union
 
 import pydantic
 from pydantic import SkipValidation
@@ -14,6 +14,7 @@ from langchain_core.utils.pydantic import (
     PydanticBaseModel,
     TBaseModel,
 )
+from langchain_core.v1.messages import AIMessage
 
 
 class PydanticOutputParser(JsonOutputParser, Generic[TBaseModel]):
@@ -43,7 +44,7 @@ class PydanticOutputParser(JsonOutputParser, Generic[TBaseModel]):
         return OutputParserException(msg, llm_output=json_string)
 
     def parse_result(
-        self, result: list[Generation], *, partial: bool = False
+        self, result: Union[list[Generation], AIMessage], *, partial: bool = False
     ) -> Optional[TBaseModel]:
         """Parse the result of an LLM call to a pydantic object.
 

libs/core/langchain_core/output_parsers/transform.py

@@ -20,6 +20,7 @@ from langchain_core.outputs import (
     GenerationChunk,
 )
 from langchain_core.runnables.config import run_in_executor
+from langchain_core.v1.messages import AIMessage, AIMessageChunk
 
 if TYPE_CHECKING:
     from collections.abc import AsyncIterator, Iterator
@@ -32,23 +33,27 @@ class BaseTransformOutputParser(BaseOutputParser[T]):
 
     def _transform(
         self,
-        input: Iterator[Union[str, BaseMessage]],  # noqa: A002
+        input: Iterator[Union[str, BaseMessage, AIMessage]],
     ) -> Iterator[T]:
         for chunk in input:
             if isinstance(chunk, BaseMessage):
                 yield self.parse_result([ChatGeneration(message=chunk)])
+            elif isinstance(chunk, AIMessage):
+                yield self.parse_result(chunk)
             else:
                 yield self.parse_result([Generation(text=chunk)])
 
     async def _atransform(
         self,
-        input: AsyncIterator[Union[str, BaseMessage]],  # noqa: A002
+        input: AsyncIterator[Union[str, BaseMessage, AIMessage]],
     ) -> AsyncIterator[T]:
         async for chunk in input:
             if isinstance(chunk, BaseMessage):
                 yield await run_in_executor(
                     None, self.parse_result, [ChatGeneration(message=chunk)]
                 )
+            elif isinstance(chunk, AIMessage):
+                yield await run_in_executor(None, self.parse_result, chunk)
             else:
                 yield await run_in_executor(
                     None, self.parse_result, [Generation(text=chunk)]
@@ -57,7 +62,7 @@ class BaseTransformOutputParser(BaseOutputParser[T]):
     @override
     def transform(
         self,
-        input: Iterator[Union[str, BaseMessage]],
+        input: Iterator[Union[str, BaseMessage, AIMessage]],
         config: Optional[RunnableConfig] = None,
         **kwargs: Any,
     ) -> Iterator[T]:
@@ -78,7 +83,7 @@ class BaseTransformOutputParser(BaseOutputParser[T]):
     @override
     async def atransform(
         self,
-        input: AsyncIterator[Union[str, BaseMessage]],
+        input: AsyncIterator[Union[str, BaseMessage, AIMessage]],
         config: Optional[RunnableConfig] = None,
         **kwargs: Any,
     ) -> AsyncIterator[T]:
@@ -125,23 +130,42 @@ class BaseCumulativeTransformOutputParser(BaseTransformOutputParser[T]):
         raise NotImplementedError
 
     @override
-    def _transform(self, input: Iterator[Union[str, BaseMessage]]) -> Iterator[Any]:
+    def _transform(
+        self, input: Iterator[Union[str, BaseMessage, AIMessage]]
+    ) -> Iterator[Any]:
         prev_parsed = None
-        acc_gen: Union[GenerationChunk, ChatGenerationChunk, None] = None
+        acc_gen: Union[GenerationChunk, ChatGenerationChunk, AIMessageChunk, None] = (
+            None
+        )
         for chunk in input:
-            chunk_gen: Union[GenerationChunk, ChatGenerationChunk]
+            chunk_gen: Union[GenerationChunk, ChatGenerationChunk, AIMessageChunk]
             if isinstance(chunk, BaseMessageChunk):
                 chunk_gen = ChatGenerationChunk(message=chunk)
             elif isinstance(chunk, BaseMessage):
                 chunk_gen = ChatGenerationChunk(
                     message=BaseMessageChunk(**chunk.model_dump())
                 )
+            elif isinstance(chunk, AIMessageChunk):
+                chunk_gen = chunk
+            elif isinstance(chunk, AIMessage):
+                chunk_gen = AIMessageChunk(
+                    content=chunk.content,
+                    id=chunk.id,
+                    name=chunk.name,
+                    lc_version=chunk.lc_version,
+                    response_metadata=chunk.response_metadata,
+                    usage_metadata=chunk.usage_metadata,
+                    parsed=chunk.parsed,
+                )
             else:
                 chunk_gen = GenerationChunk(text=chunk)
 
             acc_gen = chunk_gen if acc_gen is None else acc_gen + chunk_gen  # type: ignore[operator]
 
-            parsed = self.parse_result([acc_gen], partial=True)
+            if isinstance(acc_gen, AIMessageChunk):
+                parsed = self.parse_result(acc_gen, partial=True)
+            else:
+                parsed = self.parse_result([acc_gen], partial=True)
             if parsed is not None and parsed != prev_parsed:
                 if self.diff:
                     yield self._diff(prev_parsed, parsed)
@@ -151,24 +175,41 @@ class BaseCumulativeTransformOutputParser(BaseTransformOutputParser[T]):
 
     @override
     async def _atransform(
-        self, input: AsyncIterator[Union[str, BaseMessage]]
+        self, input: AsyncIterator[Union[str, BaseMessage, AIMessage]]
     ) -> AsyncIterator[T]:
         prev_parsed = None
-        acc_gen: Union[GenerationChunk, ChatGenerationChunk, None] = None
+        acc_gen: Union[GenerationChunk, ChatGenerationChunk, AIMessageChunk, None] = (
+            None
+        )
         async for chunk in input:
-            chunk_gen: Union[GenerationChunk, ChatGenerationChunk]
+            chunk_gen: Union[GenerationChunk, ChatGenerationChunk, AIMessageChunk]
             if isinstance(chunk, BaseMessageChunk):
                 chunk_gen = ChatGenerationChunk(message=chunk)
             elif isinstance(chunk, BaseMessage):
                 chunk_gen = ChatGenerationChunk(
                     message=BaseMessageChunk(**chunk.model_dump())
                 )
+            elif isinstance(chunk, AIMessageChunk):
+                chunk_gen = chunk
+            elif isinstance(chunk, AIMessage):
+                chunk_gen = AIMessageChunk(
+                    content=chunk.content,
+                    id=chunk.id,
+                    name=chunk.name,
+                    lc_version=chunk.lc_version,
+                    response_metadata=chunk.response_metadata,
+                    usage_metadata=chunk.usage_metadata,
+                    parsed=chunk.parsed,
+                )
             else:
                 chunk_gen = GenerationChunk(text=chunk)
 
             acc_gen = chunk_gen if acc_gen is None else acc_gen + chunk_gen  # type: ignore[operator]
 
-            parsed = await self.aparse_result([acc_gen], partial=True)
+            if isinstance(acc_gen, AIMessageChunk):
+                parsed = await self.aparse_result(acc_gen, partial=True)
+            else:
+                parsed = await self.aparse_result([acc_gen], partial=True)
             if parsed is not None and parsed != prev_parsed:
                 if self.diff:
                     yield await run_in_executor(None, self._diff, prev_parsed, parsed)

libs/core/langchain_core/output_parsers/xml.py

@@ -12,8 +12,10 @@ from typing_extensions import override
 
 from langchain_core.exceptions import OutputParserException
 from langchain_core.messages import BaseMessage
+from langchain_core.messages.utils import convert_from_v1_message
 from langchain_core.output_parsers.transform import BaseTransformOutputParser
 from langchain_core.runnables.utils import AddableDict
+from langchain_core.v1.messages import AIMessage
 
 XML_FORMAT_INSTRUCTIONS = """The output should be formatted as a XML file.
 1. Output should conform to the tags below.
@@ -105,23 +107,27 @@ class _StreamingParser:
         self.buffer = ""
         # yield all events
         try:
-            for event, elem in self.pull_parser.read_events():
-                if event == "start":
-                    # update current path
-                    self.current_path.append(elem.tag)
-                    self.current_path_has_children = False
-                elif event == "end":
-                    # remove last element from current path
-                    #
-                    self.current_path.pop()
-                    # yield element
-                    if not self.current_path_has_children:
-                        yield nested_element(self.current_path, elem)
-                    # prevent yielding of parent element
-                    if self.current_path:
-                        self.current_path_has_children = True
-                    else:
-                        self.xml_started = False
+            for raw_event in self.pull_parser.read_events():
+                if len(raw_event) <= 1:
+                    continue
+                event, elem = raw_event
+                if isinstance(elem, ET.Element):
+                    if event == "start":
+                        # update current path
+                        self.current_path.append(elem.tag)
+                        self.current_path_has_children = False
+                    elif event == "end":
+                        # remove last element from current path
+                        #
+                        self.current_path.pop()
+                        # yield element
+                        if not self.current_path_has_children:
+                            yield nested_element(self.current_path, elem)
+                        # prevent yielding of parent element
+                        if self.current_path:
+                            self.current_path_has_children = True
+                        else:
+                            self.xml_started = False
         except xml.etree.ElementTree.ParseError:
             # This might be junk at the end of the XML input.
             # Let's check whether the current path is empty.
@@ -240,21 +246,28 @@ class XMLOutputParser(BaseTransformOutputParser):
 
     @override
     def _transform(
-        self, input: Iterator[Union[str, BaseMessage]]
+        self, input: Iterator[Union[str, BaseMessage, AIMessage]]
     ) -> Iterator[AddableDict]:
         streaming_parser = _StreamingParser(self.parser)
         for chunk in input:
-            yield from streaming_parser.parse(chunk)
+            if isinstance(chunk, AIMessage):
+                yield from streaming_parser.parse(convert_from_v1_message(chunk))
+            else:
+                yield from streaming_parser.parse(chunk)
         streaming_parser.close()
 
     @override
     async def _atransform(
-        self, input: AsyncIterator[Union[str, BaseMessage]]
+        self, input: AsyncIterator[Union[str, BaseMessage, AIMessage]]
     ) -> AsyncIterator[AddableDict]:
         streaming_parser = _StreamingParser(self.parser)
         async for chunk in input:
-            for output in streaming_parser.parse(chunk):
-                yield output
+            if isinstance(chunk, AIMessage):
+                for output in streaming_parser.parse(convert_from_v1_message(chunk)):
+                    yield output
+            else:
+                for output in streaming_parser.parse(chunk):
+                    yield output
         streaming_parser.close()
 
     def _root_to_dict(self, root: ET.Element) -> dict[str, Union[str, list[Any]]]:

libs/core/langchain_core/prompt_values.py

@@ -8,17 +8,65 @@ from __future__ import annotations
 
 from abc import ABC, abstractmethod
 from collections.abc import Sequence
-from typing import Literal, cast
+from typing import Literal, Union, cast
 
-from typing_extensions import TypedDict
+from typing_extensions import TypedDict, overload
 
 from langchain_core.load.serializable import Serializable
 from langchain_core.messages import (
+    AIMessage,
     AnyMessage,
     BaseMessage,
     HumanMessage,
+    SystemMessage,
+    ToolMessage,
     get_buffer_string,
 )
+from langchain_core.messages import content_blocks as types
+from langchain_core.v1.messages import AIMessage as AIMessageV1
+from langchain_core.v1.messages import HumanMessage as HumanMessageV1
+from langchain_core.v1.messages import MessageV1, ResponseMetadata
+from langchain_core.v1.messages import SystemMessage as SystemMessageV1
+from langchain_core.v1.messages import ToolMessage as ToolMessageV1
+
+
+def _convert_to_v1(message: BaseMessage) -> MessageV1:
+    """Best-effort conversion of a V0 AIMessage to V1."""
+    if isinstance(message.content, str):
+        content: list[types.ContentBlock] = []
+        if message.content:
+            content = [{"type": "text", "text": message.content}]
+    else:
+        content = []
+        for block in message.content:
+            if isinstance(block, str):
+                content.append({"type": "text", "text": block})
+            elif isinstance(block, dict):
+                content.append(cast("types.ContentBlock", block))
+            else:
+                pass
+
+    if isinstance(message, HumanMessage):
+        return HumanMessageV1(content=content)
+    if isinstance(message, AIMessage):
+        for tool_call in message.tool_calls:
+            content.append(tool_call)
+        return AIMessageV1(
+            content=content,
+            usage_metadata=message.usage_metadata,
+            response_metadata=cast("ResponseMetadata", message.response_metadata),
+            tool_calls=message.tool_calls,
+        )
+    if isinstance(message, SystemMessage):
+        return SystemMessageV1(content=content)
+    if isinstance(message, ToolMessage):
+        return ToolMessageV1(
+            tool_call_id=message.tool_call_id,
+            content=content,
+            artifact=message.artifact,
+        )
+    error_message = f"Unsupported message type: {type(message)}"
+    raise TypeError(error_message)
 
 
 class PromptValue(Serializable, ABC):
@@ -46,8 +94,18 @@ class PromptValue(Serializable, ABC):
     def to_string(self) -> str:
         """Return prompt value as string."""
 
+    @overload
+    def to_messages(
+        self, message_version: Literal["v0"] = "v0"
+    ) -> list[BaseMessage]: ...
+
+    @overload
+    def to_messages(self, message_version: Literal["v1"]) -> list[MessageV1]: ...
+
     @abstractmethod
-    def to_messages(self) -> list[BaseMessage]:
+    def to_messages(
+        self, message_version: Literal["v0", "v1"] = "v0"
+    ) -> Union[Sequence[BaseMessage], Sequence[MessageV1]]:
         """Return prompt as a list of Messages."""
 
 
@@ -71,8 +129,20 @@ class StringPromptValue(PromptValue):
         """Return prompt as string."""
         return self.text
 
-    def to_messages(self) -> list[BaseMessage]:
+    @overload
+    def to_messages(
+        self, message_version: Literal["v0"] = "v0"
+    ) -> list[BaseMessage]: ...
+
+    @overload
+    def to_messages(self, message_version: Literal["v1"]) -> list[MessageV1]: ...
+
+    def to_messages(
+        self, message_version: Literal["v0", "v1"] = "v0"
+    ) -> Union[Sequence[BaseMessage], Sequence[MessageV1]]:
         """Return prompt as messages."""
+        if message_version == "v1":
+            return [HumanMessageV1(content=self.text)]
         return [HumanMessage(content=self.text)]
 
 
@@ -89,8 +159,24 @@ class ChatPromptValue(PromptValue):
         """Return prompt as string."""
         return get_buffer_string(self.messages)
 
-    def to_messages(self) -> list[BaseMessage]:
-        """Return prompt as a list of messages."""
+    @overload
+    def to_messages(
+        self, message_version: Literal["v0"] = "v0"
+    ) -> list[BaseMessage]: ...
+
+    @overload
+    def to_messages(self, message_version: Literal["v1"]) -> list[MessageV1]: ...
+
+    def to_messages(
+        self, message_version: Literal["v0", "v1"] = "v0"
+    ) -> Union[Sequence[BaseMessage], Sequence[MessageV1]]:
+        """Return prompt as a list of messages.
+
+        Args:
+            message_version: The output version, either "v0" (default) or "v1".
+        """
+        if message_version == "v1":
+            return [_convert_to_v1(m) for m in self.messages]
         return list(self.messages)
 
     @classmethod
@@ -125,8 +211,26 @@ class ImagePromptValue(PromptValue):
         """Return prompt (image URL) as string."""
         return self.image_url["url"]
 
-    def to_messages(self) -> list[BaseMessage]:
+    @overload
+    def to_messages(
+        self, message_version: Literal["v0"] = "v0"
+    ) -> list[BaseMessage]: ...
+
+    @overload
+    def to_messages(self, message_version: Literal["v1"]) -> list[MessageV1]: ...
+
+    def to_messages(
+        self, message_version: Literal["v0", "v1"] = "v0"
+    ) -> Union[Sequence[BaseMessage], Sequence[MessageV1]]:
         """Return prompt (image URL) as messages."""
+        if message_version == "v1":
+            block: types.ImageContentBlock = {
+                "type": "image",
+                "url": self.image_url["url"],
+            }
+            if "detail" in self.image_url:
+                block["detail"] = self.image_url["detail"]
+            return [HumanMessageV1(content=[block])]
         return [HumanMessage(content=[cast("dict", self.image_url)])]
 
 

libs/core/langchain_core/runnables/config.py

@@ -402,7 +402,7 @@ def call_func_with_variable_args(
         Callable[[Input, CallbackManagerForChainRun], Output],
         Callable[[Input, CallbackManagerForChainRun, RunnableConfig], Output],
     ],
-    input: Input,  # noqa: A002
+    input: Input,
     config: RunnableConfig,
     run_manager: Optional[CallbackManagerForChainRun] = None,
     **kwargs: Any,
@@ -439,7 +439,7 @@ def acall_func_with_variable_args(
             Awaitable[Output],
         ],
     ],
-    input: Input,  # noqa: A002
+    input: Input,
     config: RunnableConfig,
     run_manager: Optional[AsyncCallbackManagerForChainRun] = None,
     **kwargs: Any,

libs/core/langchain_core/runnables/fallbacks.py

@@ -5,7 +5,7 @@ import inspect
 import typing
 from collections.abc import AsyncIterator, Iterator, Sequence
 from functools import wraps
-from typing import TYPE_CHECKING, Any, Optional, Union
+from typing import TYPE_CHECKING, Any, Optional, Union, cast
 
 from pydantic import BaseModel, ConfigDict
 from typing_extensions import override
@@ -397,7 +397,7 @@ class RunnableWithFallbacks(RunnableSerializable[Input, Output]):
             )
         )
 
-        to_return = {}
+        to_return: dict[int, Union[Output, BaseException]] = {}
         run_again = dict(enumerate(inputs))
         handled_exceptions: dict[int, BaseException] = {}
         first_to_raise = None
@@ -447,7 +447,7 @@ class RunnableWithFallbacks(RunnableSerializable[Input, Output]):
         if not return_exceptions and sorted_handled_exceptions:
             raise sorted_handled_exceptions[0][1]
         to_return.update(handled_exceptions)
-        return [output for _, output in sorted(to_return.items())]  # type: ignore[misc]
+        return [cast("Output", output) for _, output in sorted(to_return.items())]
 
     @override
     def stream(
@@ -569,7 +569,7 @@ class RunnableWithFallbacks(RunnableSerializable[Input, Output]):
             async for chunk in stream:
                 yield chunk
                 try:
-                    output = output + chunk
+                    output = output + chunk  # type: ignore[operator]
                 except TypeError:
                     output = None
         except BaseException as e:

libs/core/langchain_core/runnables/graph.py

@@ -114,7 +114,7 @@ class Node(NamedTuple):
     def copy(
         self,
         *,
-        id: Optional[str] = None,  # noqa: A002
+        id: Optional[str] = None,
         name: Optional[str] = None,
     ) -> Node:
         """Return a copy of the node with optional new id and name.
@@ -187,7 +187,7 @@ class MermaidDrawMethod(Enum):
 
 
 def node_data_str(
-    id: str,  # noqa: A002
+    id: str,
     data: Union[type[BaseModel], RunnableType, None],
 ) -> str:
     """Convert the data of a node to a string.
@@ -328,7 +328,7 @@ class Graph:
     def add_node(
         self,
         data: Union[type[BaseModel], RunnableType, None],
-        id: Optional[str] = None,  # noqa: A002
+        id: Optional[str] = None,
         *,
         metadata: Optional[dict[str, Any]] = None,
     ) -> Node:

libs/core/langchain_core/tools/base.py

@@ -68,13 +68,21 @@ from langchain_core.utils.pydantic import (
     is_pydantic_v1_subclass,
     is_pydantic_v2_subclass,
 )
+from langchain_core.v1.messages import ToolMessage as ToolMessageV1
 
 if TYPE_CHECKING:
     import uuid
     from collections.abc import Sequence
 
 FILTERED_ARGS = ("run_manager", "callbacks")
-TOOL_MESSAGE_BLOCK_TYPES = ("text", "image_url", "image", "json", "search_result")
+TOOL_MESSAGE_BLOCK_TYPES = (
+    "text",
+    "image_url",
+    "image",
+    "json",
+    "search_result",
+    "custom_tool_call_output",
+)
 
 
 class SchemaAnnotationError(TypeError):
@@ -498,6 +506,15 @@ class ChildTool(BaseTool):
     two-tuple corresponding to the (content, artifact) of a ToolMessage.
     """
 
+    message_version: Literal["v0", "v1"] = "v0"
+    """Version of ToolMessage to return given
+    :class:`~langchain_core.messages.content_blocks.ToolCall` input.
+
+    If ``"v0"``, output will be a v0 :class:`~langchain_core.messages.tool.ToolMessage`.
+    If ``"v1"``, output will be a v1 :class:`~langchain_core.v1.messages.ToolMessage`.
+
+    """
+
     def __init__(self, **kwargs: Any) -> None:
         """Initialize the tool."""
         if (
@@ -835,7 +852,7 @@ class ChildTool(BaseTool):
 
         content = None
         artifact = None
-        status = "success"
+        status: Literal["success", "error"] = "success"
         error_to_raise: Union[Exception, KeyboardInterrupt, None] = None
         try:
             child_config = patch_config(config, callbacks=run_manager.get_child())
@@ -879,7 +896,14 @@ class ChildTool(BaseTool):
         if error_to_raise:
             run_manager.on_tool_error(error_to_raise)
             raise error_to_raise
-        output = _format_output(content, artifact, tool_call_id, self.name, status)
+        output = _format_output(
+            content,
+            artifact,
+            tool_call_id,
+            self.name,
+            status,
+            message_version=self.message_version,
+        )
         run_manager.on_tool_end(output, color=color, name=self.name, **kwargs)
         return output
 
@@ -945,7 +969,7 @@ class ChildTool(BaseTool):
         )
         content = None
         artifact = None
-        status = "success"
+        status: Literal["success", "error"] = "success"
         error_to_raise: Optional[Union[Exception, KeyboardInterrupt]] = None
         try:
             tool_args, tool_kwargs = self._to_args_and_kwargs(tool_input, tool_call_id)
@@ -993,7 +1017,14 @@ class ChildTool(BaseTool):
             await run_manager.on_tool_error(error_to_raise)
             raise error_to_raise
 
-        output = _format_output(content, artifact, tool_call_id, self.name, status)
+        output = _format_output(
+            content,
+            artifact,
+            tool_call_id,
+            self.name,
+            status,
+            message_version=self.message_version,
+        )
         await run_manager.on_tool_end(output, color=color, name=self.name, **kwargs)
         return output
 
@@ -1131,7 +1162,9 @@ def _format_output(
     artifact: Any,
     tool_call_id: Optional[str],
     name: str,
-    status: str,
+    status: Literal["success", "error"],
+    *,
+    message_version: Literal["v0", "v1"] = "v0",
 ) -> Union[ToolOutputMixin, Any]:
     """Format tool output as a ToolMessage if appropriate.
 
@@ -1141,6 +1174,7 @@ def _format_output(
         tool_call_id: The ID of the tool call.
         name: The name of the tool.
         status: The execution status.
+        message_version: The version of the ToolMessage to return.
 
     Returns:
         The formatted output, either as a ToolMessage or the original content.
@@ -1149,7 +1183,15 @@ def _format_output(
         return content
     if not _is_message_content_type(content):
         content = _stringify(content)
-    return ToolMessage(
+    if message_version == "v0":
+        return ToolMessage(
+            content,
+            artifact=artifact,
+            tool_call_id=tool_call_id,
+            name=name,
+            status=status,
+        )
+    return ToolMessageV1(
         content,
         artifact=artifact,
         tool_call_id=tool_call_id,

libs/core/langchain_core/tools/convert.py

@@ -22,6 +22,7 @@ def tool(
     response_format: Literal["content", "content_and_artifact"] = "content",
     parse_docstring: bool = False,
     error_on_invalid_docstring: bool = True,
+    message_version: Literal["v0", "v1"] = "v0",
 ) -> Callable[[Union[Callable, Runnable]], BaseTool]: ...
 
 
@@ -37,6 +38,7 @@ def tool(
     response_format: Literal["content", "content_and_artifact"] = "content",
     parse_docstring: bool = False,
     error_on_invalid_docstring: bool = True,
+    message_version: Literal["v0", "v1"] = "v0",
 ) -> BaseTool: ...
 
 
@@ -51,6 +53,7 @@ def tool(
     response_format: Literal["content", "content_and_artifact"] = "content",
     parse_docstring: bool = False,
     error_on_invalid_docstring: bool = True,
+    message_version: Literal["v0", "v1"] = "v0",
 ) -> BaseTool: ...
 
 
@@ -65,6 +68,7 @@ def tool(
     response_format: Literal["content", "content_and_artifact"] = "content",
     parse_docstring: bool = False,
     error_on_invalid_docstring: bool = True,
+    message_version: Literal["v0", "v1"] = "v0",
 ) -> Callable[[Union[Callable, Runnable]], BaseTool]: ...
 
 
@@ -79,6 +83,7 @@ def tool(
     response_format: Literal["content", "content_and_artifact"] = "content",
     parse_docstring: bool = False,
     error_on_invalid_docstring: bool = True,
+    message_version: Literal["v0", "v1"] = "v0",
 ) -> Union[
     BaseTool,
     Callable[[Union[Callable, Runnable]], BaseTool],
@@ -118,6 +123,11 @@ def tool(
         error_on_invalid_docstring: if ``parse_docstring`` is provided, configure
             whether to raise ValueError on invalid Google Style docstrings.
             Defaults to True.
+        message_version: Version of ToolMessage to return given
+            :class:`~langchain_core.messages.content_blocks.ToolCall` input.
+
+            If ``"v0"``, output will be a v0 :class:`~langchain_core.messages.tool.ToolMessage`.
+            If ``"v1"``, output will be a v1 :class:`~langchain_core.v1.messages.ToolMessage`.
 
     Returns:
         The tool.
@@ -216,7 +226,7 @@ def tool(
                 \"\"\"
                 return bar
 
-    """  # noqa: D214, D410, D411
+    """  # noqa: D214, D410, D411, E501
 
     def _create_tool_factory(
         tool_name: str,
@@ -274,6 +284,7 @@ def tool(
                     response_format=response_format,
                     parse_docstring=parse_docstring,
                     error_on_invalid_docstring=error_on_invalid_docstring,
+                    message_version=message_version,
                 )
             # If someone doesn't want a schema applied, we must treat it as
             # a simple string->string function
@@ -290,6 +301,7 @@ def tool(
                 return_direct=return_direct,
                 coroutine=coroutine,
                 response_format=response_format,
+                message_version=message_version,
             )
 
         return _tool_factory
@@ -383,6 +395,7 @@ def convert_runnable_to_tool(
     name: Optional[str] = None,
     description: Optional[str] = None,
     arg_types: Optional[dict[str, type]] = None,
+    message_version: Literal["v0", "v1"] = "v0",
 ) -> BaseTool:
     """Convert a Runnable into a BaseTool.
 
@@ -392,10 +405,15 @@ def convert_runnable_to_tool(
         name: The name of the tool. Defaults to None.
         description: The description of the tool. Defaults to None.
         arg_types: The types of the arguments. Defaults to None.
+        message_version: Version of ToolMessage to return given
+            :class:`~langchain_core.messages.content_blocks.ToolCall` input.
+
+            If ``"v0"``, output will be a v0 :class:`~langchain_core.messages.tool.ToolMessage`.
+            If ``"v1"``, output will be a v1 :class:`~langchain_core.v1.messages.ToolMessage`.
 
     Returns:
         The tool.
-    """
+    """  # noqa: E501
     if args_schema:
         runnable = runnable.with_types(input_type=args_schema)
     description = description or _get_description_from_runnable(runnable)
@@ -408,6 +426,7 @@ def convert_runnable_to_tool(
             func=runnable.invoke,
             coroutine=runnable.ainvoke,
             description=description,
+            message_version=message_version,
         )
 
     async def ainvoke_wrapper(
@@ -435,4 +454,5 @@ def convert_runnable_to_tool(
         coroutine=ainvoke_wrapper,
         description=description,
         args_schema=args_schema,
+        message_version=message_version,
     )

libs/core/langchain_core/tools/retriever.py

@@ -72,6 +72,7 @@ def create_retriever_tool(
     document_prompt: Optional[BasePromptTemplate] = None,
     document_separator: str = "\n\n",
     response_format: Literal["content", "content_and_artifact"] = "content",
+    message_version: Literal["v0", "v1"] = "v1",
 ) -> Tool:
     r"""Create a tool to do retrieval of documents.
 
@@ -88,10 +89,15 @@ def create_retriever_tool(
             "content_and_artifact" then the output is expected to be a two-tuple
             corresponding to the (content, artifact) of a ToolMessage (artifact
             being a list of documents in this case). Defaults to "content".
+        message_version: Version of ToolMessage to return given
+            :class:`~langchain_core.messages.content_blocks.ToolCall` input.
+
+            If ``"v0"``, output will be a v0 :class:`~langchain_core.messages.tool.ToolMessage`.
+            If ``"v1"``, output will be a v1 :class:`~langchain_core.v1.messages.ToolMessage`.
 
     Returns:
         Tool class to pass to an agent.
-    """
+    """  # noqa: E501
     document_prompt = document_prompt or PromptTemplate.from_template("{page_content}")
     func = partial(
         _get_relevant_documents,
@@ -114,4 +120,5 @@ def create_retriever_tool(
         coroutine=afunc,
         args_schema=RetrieverInput,
         response_format=response_format,
+        message_version=message_version,
     )

libs/core/langchain_core/tools/structured.py

@@ -129,6 +129,7 @@ class StructuredTool(BaseTool):
         response_format: Literal["content", "content_and_artifact"] = "content",
         parse_docstring: bool = False,
         error_on_invalid_docstring: bool = False,
+        message_version: Literal["v0", "v1"] = "v0",
         **kwargs: Any,
     ) -> StructuredTool:
         """Create tool from a given function.
@@ -157,6 +158,12 @@ class StructuredTool(BaseTool):
             error_on_invalid_docstring: if ``parse_docstring`` is provided, configure
                 whether to raise ValueError on invalid Google Style docstrings.
                 Defaults to False.
+            message_version: Version of ToolMessage to return given
+                :class:`~langchain_core.messages.content_blocks.ToolCall` input.
+
+                If ``"v0"``, output will be a v0 :class:`~langchain_core.messages.tool.ToolMessage`.
+                If ``"v1"``, output will be a v1 :class:`~langchain_core.v1.messages.ToolMessage`.
+
             kwargs: Additional arguments to pass to the tool
 
         Returns:
@@ -175,7 +182,7 @@ class StructuredTool(BaseTool):
                 tool = StructuredTool.from_function(add)
                 tool.run(1, 2) # 3
 
-        """
+        """  # noqa: E501
         if func is not None:
             source_function = func
         elif coroutine is not None:
@@ -232,6 +239,7 @@ class StructuredTool(BaseTool):
             description=description_,
             return_direct=return_direct,
             response_format=response_format,
+            message_version=message_version,
             **kwargs,
         )
 

libs/core/langchain_core/tracers/base.py

@@ -17,6 +17,7 @@ from typing_extensions import override
 from langchain_core.callbacks.base import AsyncCallbackHandler, BaseCallbackHandler
 from langchain_core.exceptions import TracerException  # noqa: F401
 from langchain_core.tracers.core import _TracerCore
+from langchain_core.v1.messages import AIMessage, AIMessageChunk, MessageV1
 
 if TYPE_CHECKING:
     from collections.abc import Sequence
@@ -54,7 +55,7 @@ class BaseTracer(_TracerCore, BaseCallbackHandler, ABC):
     def on_chat_model_start(
         self,
         serialized: dict[str, Any],
-        messages: list[list[BaseMessage]],
+        messages: Union[list[list[BaseMessage]], list[MessageV1]],
         *,
         run_id: UUID,
         tags: Optional[list[str]] = None,
@@ -138,7 +139,9 @@ class BaseTracer(_TracerCore, BaseCallbackHandler, ABC):
         self,
         token: str,
         *,
-        chunk: Optional[Union[GenerationChunk, ChatGenerationChunk]] = None,
+        chunk: Optional[
+            Union[GenerationChunk, ChatGenerationChunk, AIMessageChunk]
+        ] = None,
         run_id: UUID,
         parent_run_id: Optional[UUID] = None,
         **kwargs: Any,
@@ -190,7 +193,9 @@ class BaseTracer(_TracerCore, BaseCallbackHandler, ABC):
         )
 
     @override
-    def on_llm_end(self, response: LLMResult, *, run_id: UUID, **kwargs: Any) -> Run:
+    def on_llm_end(
+        self, response: Union[LLMResult, AIMessage], *, run_id: UUID, **kwargs: Any
+    ) -> Run:
         """End a trace for an LLM run.
 
         Args:
@@ -562,7 +567,7 @@ class AsyncBaseTracer(_TracerCore, AsyncCallbackHandler, ABC):
     async def on_chat_model_start(
         self,
         serialized: dict[str, Any],
-        messages: list[list[BaseMessage]],
+        messages: Union[list[list[BaseMessage]], list[MessageV1]],
         *,
         run_id: UUID,
         parent_run_id: Optional[UUID] = None,
@@ -617,7 +622,9 @@ class AsyncBaseTracer(_TracerCore, AsyncCallbackHandler, ABC):
         self,
         token: str,
         *,
-        chunk: Optional[Union[GenerationChunk, ChatGenerationChunk]] = None,
+        chunk: Optional[
+            Union[GenerationChunk, ChatGenerationChunk, AIMessageChunk]
+        ] = None,
         run_id: UUID,
         parent_run_id: Optional[UUID] = None,
         **kwargs: Any,
@@ -646,7 +653,7 @@ class AsyncBaseTracer(_TracerCore, AsyncCallbackHandler, ABC):
     @override
     async def on_llm_end(
         self,
-        response: LLMResult,
+        response: Union[LLMResult, AIMessage],
         *,
         run_id: UUID,
         parent_run_id: Optional[UUID] = None,
@@ -882,7 +889,7 @@ class AsyncBaseTracer(_TracerCore, AsyncCallbackHandler, ABC):
         self,
         run: Run,
         token: str,
-        chunk: Optional[Union[GenerationChunk, ChatGenerationChunk]],
+        chunk: Optional[Union[GenerationChunk, ChatGenerationChunk, AIMessageChunk]],
     ) -> None:
         """Process new LLM token."""
 

libs/core/langchain_core/tracers/core.py

@@ -18,6 +18,7 @@ from typing import (
 
 from langchain_core.exceptions import TracerException
 from langchain_core.load import dumpd
+from langchain_core.messages.utils import convert_from_v1_message
 from langchain_core.outputs import (
     ChatGeneration,
     ChatGenerationChunk,
@@ -25,6 +26,12 @@ from langchain_core.outputs import (
     LLMResult,
 )
 from langchain_core.tracers.schemas import Run
+from langchain_core.v1.messages import (
+    AIMessage,
+    AIMessageChunk,
+    MessageV1,
+    MessageV1Types,
+)
 
 if TYPE_CHECKING:
     from collections.abc import Coroutine, Sequence
@@ -156,7 +163,7 @@ class _TracerCore(ABC):
     def _create_chat_model_run(
         self,
         serialized: dict[str, Any],
-        messages: list[list[BaseMessage]],
+        messages: Union[list[list[BaseMessage]], list[MessageV1]],
         run_id: UUID,
         tags: Optional[list[str]] = None,
         parent_run_id: Optional[UUID] = None,
@@ -181,6 +188,12 @@ class _TracerCore(ABC):
         start_time = datetime.now(timezone.utc)
         if metadata:
             kwargs.update({"metadata": metadata})
+        if isinstance(messages[0], MessageV1Types):
+            # Convert from v1 messages to BaseMessage
+            messages = [
+                [convert_from_v1_message(msg) for msg in messages]  # type: ignore[arg-type]
+            ]
+        messages = cast("list[list[BaseMessage]]", messages)
         return Run(
             id=run_id,
             parent_run_id=parent_run_id,
@@ -230,7 +243,9 @@ class _TracerCore(ABC):
         self,
         token: str,
         run_id: UUID,
-        chunk: Optional[Union[GenerationChunk, ChatGenerationChunk]] = None,
+        chunk: Optional[
+            Union[GenerationChunk, ChatGenerationChunk, AIMessageChunk]
+        ] = None,
         parent_run_id: Optional[UUID] = None,  # noqa: ARG002
     ) -> Run:
         """Append token event to LLM run and return the run."""
@@ -276,7 +291,15 @@ class _TracerCore(ABC):
         )
         return llm_run
 
-    def _complete_llm_run(self, response: LLMResult, run_id: UUID) -> Run:
+    def _complete_llm_run(
+        self, response: Union[LLMResult, AIMessage], run_id: UUID
+    ) -> Run:
+        if isinstance(response, AIMessage):
+            response = LLMResult(
+                generations=[
+                    [ChatGeneration(message=convert_from_v1_message(response))]
+                ]
+            )
         llm_run = self._get_run(run_id, run_type={"llm", "chat_model"})
         if getattr(llm_run, "outputs", None) is None:
             llm_run.outputs = {}
@@ -558,7 +581,7 @@ class _TracerCore(ABC):
         self,
         run: Run,  # noqa: ARG002
         token: str,  # noqa: ARG002
-        chunk: Optional[Union[GenerationChunk, ChatGenerationChunk]],  # noqa: ARG002
+        chunk: Optional[Union[GenerationChunk, ChatGenerationChunk, AIMessageChunk]],  # noqa: ARG002
     ) -> Union[None, Coroutine[Any, Any, None]]:
         """Process new LLM token."""
         return None

libs/core/langchain_core/tracers/event_stream.py

@@ -38,6 +38,7 @@ from langchain_core.runnables.utils import (
 from langchain_core.tracers._streaming import _StreamingCallbackHandler
 from langchain_core.tracers.memory_stream import _MemoryStream
 from langchain_core.utils.aiter import aclosing, py_anext
+from langchain_core.v1.messages import MessageV1
 
 if TYPE_CHECKING:
     from collections.abc import AsyncIterator, Iterator, Sequence
@@ -45,6 +46,8 @@ if TYPE_CHECKING:
     from langchain_core.documents import Document
     from langchain_core.runnables import Runnable, RunnableConfig
     from langchain_core.tracers.log_stream import LogEntry
+    from langchain_core.v1.messages import AIMessage as AIMessageV1
+    from langchain_core.v1.messages import AIMessageChunk as AIMessageChunkV1
 
 logger = logging.getLogger(__name__)
 
@@ -297,7 +300,7 @@ class _AstreamEventsCallbackHandler(AsyncCallbackHandler, _StreamingCallbackHand
     async def on_chat_model_start(
         self,
         serialized: dict[str, Any],
-        messages: list[list[BaseMessage]],
+        messages: Union[list[list[BaseMessage]], list[MessageV1]],
         *,
         run_id: UUID,
         tags: Optional[list[str]] = None,
@@ -307,6 +310,8 @@ class _AstreamEventsCallbackHandler(AsyncCallbackHandler, _StreamingCallbackHand
         **kwargs: Any,
     ) -> None:
         """Start a trace for an LLM run."""
+        # below cast is because type is converted in handle_event
+        messages = cast("list[list[BaseMessage]]", messages)
         name_ = _assign_name(name, serialized)
         run_type = "chat_model"
 
@@ -407,13 +412,18 @@ class _AstreamEventsCallbackHandler(AsyncCallbackHandler, _StreamingCallbackHand
         self,
         token: str,
         *,
-        chunk: Optional[Union[GenerationChunk, ChatGenerationChunk]] = None,
+        chunk: Optional[
+            Union[GenerationChunk, ChatGenerationChunk, AIMessageChunkV1]
+        ] = None,
         run_id: UUID,
         parent_run_id: Optional[UUID] = None,
         **kwargs: Any,
     ) -> None:
         """Run on new LLM token. Only available when streaming is enabled."""
         run_info = self.run_map.get(run_id)
+        chunk = cast(
+            "Optional[Union[GenerationChunk, ChatGenerationChunk]]", chunk
+        )  # converted in handle_event
         chunk_: Union[GenerationChunk, BaseMessageChunk]
 
         if run_info is None:
@@ -456,9 +466,10 @@ class _AstreamEventsCallbackHandler(AsyncCallbackHandler, _StreamingCallbackHand
 
     @override
     async def on_llm_end(
-        self, response: LLMResult, *, run_id: UUID, **kwargs: Any
+        self, response: Union[LLMResult, AIMessageV1], *, run_id: UUID, **kwargs: Any
     ) -> None:
         """End a trace for an LLM run."""
+        response = cast("LLMResult", response)  # converted in handle_event
         run_info = self.run_map.pop(run_id)
         inputs_ = run_info["inputs"]
 

libs/core/langchain_core/tracers/langchain.py

@@ -5,7 +5,7 @@ from __future__ import annotations
 import logging
 from concurrent.futures import ThreadPoolExecutor
 from datetime import datetime, timezone
-from typing import TYPE_CHECKING, Any, Optional, Union
+from typing import TYPE_CHECKING, Any, Optional, Union, cast
 from uuid import UUID
 
 from langsmith import Client
@@ -21,12 +21,15 @@ from typing_extensions import override
 
 from langchain_core.env import get_runtime_environment
 from langchain_core.load import dumpd
+from langchain_core.messages.utils import convert_from_v1_message
 from langchain_core.tracers.base import BaseTracer
 from langchain_core.tracers.schemas import Run
+from langchain_core.v1.messages import MessageV1Types
 
 if TYPE_CHECKING:
     from langchain_core.messages import BaseMessage
     from langchain_core.outputs import ChatGenerationChunk, GenerationChunk
+    from langchain_core.v1.messages import AIMessageChunk, MessageV1
 
 logger = logging.getLogger(__name__)
 _LOGGED = set()
@@ -113,7 +116,7 @@ class LangChainTracer(BaseTracer):
     def on_chat_model_start(
         self,
         serialized: dict[str, Any],
-        messages: list[list[BaseMessage]],
+        messages: Union[list[list[BaseMessage]], list[MessageV1]],
         *,
         run_id: UUID,
         tags: Optional[list[str]] = None,
@@ -140,6 +143,12 @@ class LangChainTracer(BaseTracer):
         start_time = datetime.now(timezone.utc)
         if metadata:
             kwargs.update({"metadata": metadata})
+        if isinstance(messages[0], MessageV1Types):
+            # Convert from v1 messages to BaseMessage
+            messages = [
+                [convert_from_v1_message(msg) for msg in messages]  # type: ignore[arg-type]
+            ]
+        messages = cast("list[list[BaseMessage]]", messages)
         chat_model_run = Run(
             id=run_id,
             parent_run_id=parent_run_id,
@@ -232,7 +241,9 @@ class LangChainTracer(BaseTracer):
         self,
         token: str,
         run_id: UUID,
-        chunk: Optional[Union[GenerationChunk, ChatGenerationChunk]] = None,
+        chunk: Optional[
+            Union[GenerationChunk, ChatGenerationChunk, AIMessageChunk]
+        ] = None,
         parent_run_id: Optional[UUID] = None,
     ) -> Run:
         """Append token event to LLM run and return the run."""

libs/core/langchain_core/tracers/log_stream.py

@@ -34,6 +34,7 @@ if TYPE_CHECKING:
 
     from langchain_core.runnables.utils import Input, Output
     from langchain_core.tracers.schemas import Run
+    from langchain_core.v1.messages import AIMessageChunk
 
 
 class LogEntry(TypedDict):
@@ -176,7 +177,7 @@ class RunLog(RunLogPatch):
         # Then compare that the ops are the same
         return super().__eq__(other)
 
-    __hash__ = None  # type: ignore[assignment]
+    __hash__ = None
 
 
 T = TypeVar("T")
@@ -485,7 +486,7 @@ class LogStreamCallbackHandler(BaseTracer, _StreamingCallbackHandler):
         self,
         run: Run,
         token: str,
-        chunk: Optional[Union[GenerationChunk, ChatGenerationChunk]],
+        chunk: Optional[Union[GenerationChunk, ChatGenerationChunk, AIMessageChunk]],
     ) -> None:
         """Process new LLM token."""
         index = self._key_map_by_run_id.get(run.id)