ollama: update tests, docs (#31736)

- docs: for the Ollama notebooks, improve the specificity of some links,
add `homebrew` install info, update some wording
- tests: reduce number of local models needed to run in half from 4 → 2
(shedding 8gb of required installs)
- bump deps (non-breaking) in anticipation of upcoming "thinking" PR

libs/partners/ollama/README.md

@@ -12,10 +12,7 @@ For the package to work, you will need to install and run the Ollama server loca
 
 To run integration tests (`make integration_tests`), you will need the following models installed in your Ollama server:
 
-- `llama3`
-- `llama3:latest`
-- `lamma3.1`
-- `gemma3:4b`
+- `llama3.1`
 - `deepseek-r1:1.5b`
 
 Install these models by running:
@@ -24,35 +21,35 @@ Install these models by running:
 ollama pull <name-of-model>
 ```
 
-## Chat Models
+## [Chat Models](https://python.langchain.com/api_reference/ollama/chat_models/langchain_ollama.chat_models.ChatOllama.html#chatollama)
 
 `ChatOllama` class exposes chat models from Ollama.
 
 ```python
 from langchain_ollama import ChatOllama
 
-llm = ChatOllama(model="llama3-groq-tool-use")
+llm = ChatOllama(model="llama3.1")
 llm.invoke("Sing a ballad of LangChain.")
 ```
 
-## Embeddings
+## [Embeddings](https://python.langchain.com/api_reference/ollama/embeddings/langchain_ollama.embeddings.OllamaEmbeddings.html#ollamaembeddings)
 
 `OllamaEmbeddings` class exposes embeddings from Ollama.
 
 ```python
 from langchain_ollama import OllamaEmbeddings
 
-embeddings = OllamaEmbeddings(model="llama3")
+embeddings = OllamaEmbeddings(model="llama3.1")
 embeddings.embed_query("What is the meaning of life?")
 ```
 
-## LLMs
+## [LLMs](https://python.langchain.com/api_reference/ollama/llms/langchain_ollama.llms.OllamaLLM.html#ollamallm)
 
-`OllamaLLM` class exposes LLMs from Ollama.
+`OllamaLLM` class exposes traditional LLMs from Ollama.
 
 ```python
 from langchain_ollama import OllamaLLM
 
-llm = OllamaLLM(model="llama3")
+llm = OllamaLLM(model="llama3.1")
 llm.invoke("The meaning of life is")
 ```

libs/partners/ollama/pyproject.toml

@@ -7,8 +7,8 @@ authors = []
 license = { text = "MIT" }
 requires-python = ">=3.9"
 dependencies = [
-    "ollama>=0.4.8,<1.0.0",
-    "langchain-core<1.0.0,>=0.3.60",
+    "ollama>=0.5.1,<1.0.0",
+    "langchain-core<1.0.0,>=0.3.66",
 ]
 name = "langchain-ollama"
 version = "0.3.3"

libs/partners/ollama/tests/integration_tests/chat_models/test_chat_models.py

@@ -8,6 +8,8 @@ from typing_extensions import TypedDict
 
 from langchain_ollama import ChatOllama
 
+DEFAULT_MODEL_NAME = "llama3.1"
+
 
 @pytest.mark.parametrize(("method"), [("function_calling"), ("json_schema")])
 def test_structured_output(method: str) -> None:
@@ -19,7 +21,7 @@ def test_structured_output(method: str) -> None:
         setup: str = Field(description="question to set up a joke")
         punchline: str = Field(description="answer to resolve the joke")
 
-    llm = ChatOllama(model="llama3.1", temperature=0)
+    llm = ChatOllama(model=DEFAULT_MODEL_NAME, temperature=0)
     query = "Tell me a joke about cats."
 
     # Pydantic
@@ -38,7 +40,7 @@ def test_structured_output(method: str) -> None:
 
     for chunk in structured_llm.stream(query):
         assert isinstance(chunk, dict)
-    assert isinstance(chunk, dict)  # for mypy
+    assert isinstance(chunk, dict)
     assert set(chunk.keys()) == {"setup", "punchline"}
 
     # Typed Dict
@@ -55,11 +57,11 @@ def test_structured_output(method: str) -> None:
 
     for chunk in structured_llm.stream(query):
         assert isinstance(chunk, dict)
-    assert isinstance(chunk, dict)  # for mypy
+    assert isinstance(chunk, dict)
     assert set(chunk.keys()) == {"setup", "punchline"}
 
 
-@pytest.mark.parametrize(("model"), [("llama3.1")])
+@pytest.mark.parametrize(("model"), [(DEFAULT_MODEL_NAME)])
 def test_structured_output_deeply_nested(model: str) -> None:
     """Test to verify structured output with a nested objects."""
     llm = ChatOllama(model=model, temperature=0)
@@ -80,7 +82,7 @@ def test_structured_output_deeply_nested(model: str) -> None:
 
         people: list[Person]
 
-    chat = llm.with_structured_output(Data)  # type: ignore[arg-type]
+    chat = llm.with_structured_output(Data)
     text = (
         "Alan Smith is 6 feet tall and has blond hair."
         "Alan Poe is 3 feet tall and has grey hair."

libs/partners/ollama/tests/integration_tests/chat_models/test_chat_models_standard.py

@@ -1,9 +1,13 @@
 """Test chat model integration using standard integration tests."""
 
+import pytest
+from langchain_core.language_models import BaseChatModel
 from langchain_tests.integration_tests import ChatModelIntegrationTests
 
 from langchain_ollama.chat_models import ChatOllama
 
+DEFAULT_MODEL_NAME = "llama3.1"
+
 
 class TestChatOllama(ChatModelIntegrationTests):
     @property
@@ -12,7 +16,7 @@ class TestChatOllama(ChatModelIntegrationTests):
 
     @property
     def chat_model_params(self) -> dict:
-        return {"model": "llama3.1"}
+        return {"model": DEFAULT_MODEL_NAME}
 
     @property
     def supports_json_mode(self) -> bool:
@@ -20,23 +24,26 @@ class TestChatOllama(ChatModelIntegrationTests):
 
     @property
     def has_tool_choice(self) -> bool:
-        return False
+        return False  # TODO: update after Ollama implements
 
+    @property
+    def supports_image_inputs(self) -> bool:
+        return True
 
-def test_image_model() -> None:
-    class ImageModelTests(ChatModelIntegrationTests):
-        @property
-        def chat_model_class(self) -> type[ChatOllama]:
-            return ChatOllama
+    @pytest.mark.xfail(
+        reason=(
+            "Will sometime encounter AssertionErrors where tool responses are "
+            "`'3'` instead of `3`"
+        )
+    )
+    def test_tool_calling(self, model: BaseChatModel) -> None:
+        super().test_tool_calling(model)
 
-        @property
-        def chat_model_params(self) -> dict:
-            return {"model": "gemma3:4b"}
-
-        @property
-        def supports_image_inputs(self) -> bool:
-            return True
-
-    test_instance = ImageModelTests()
-    model = test_instance.chat_model_class(**test_instance.chat_model_params)
-    ImageModelTests().test_image_inputs(model)
+    @pytest.mark.xfail(
+        reason=(
+            "Will sometime encounter AssertionErrors where tool responses are "
+            "`'3'` instead of `3`"
+        )
+    )
+    async def test_tool_calling_async(self, model: BaseChatModel) -> None:
+        await super().test_tool_calling_async(model)

libs/partners/ollama/tests/integration_tests/test_embeddings.py

@@ -4,6 +4,8 @@ from langchain_tests.integration_tests import EmbeddingsIntegrationTests
 
 from langchain_ollama.embeddings import OllamaEmbeddings
 
+MODEL_NAME = "llama3.1"
+
 
 class TestOllamaEmbeddings(EmbeddingsIntegrationTests):
     @property
@@ -12,4 +14,4 @@ class TestOllamaEmbeddings(EmbeddingsIntegrationTests):
 
     @property
     def embedding_model_params(self) -> dict:
-        return {"model": "llama3:latest"}
+        return {"model": MODEL_NAME}

libs/partners/ollama/tests/integration_tests/test_llms.py

@@ -1,8 +1,10 @@
 """Test OllamaLLM llm."""
 
+from langchain_core.runnables import RunnableConfig
+
 from langchain_ollama.llms import OllamaLLM
 
-MODEL_NAME = "llama3"
+MODEL_NAME = "llama3.1"
 
 
 def test_stream() -> None:
@@ -54,13 +56,12 @@ async def test_ainvoke() -> None:
     """Test invoke tokens from OllamaLLM."""
     llm = OllamaLLM(model=MODEL_NAME)
 
-    result = await llm.ainvoke("I'm Pickle Rick", config={"tags": ["foo"]})
+    result = await llm.ainvoke("I'm Pickle Rick", config=RunnableConfig(tags=["foo"]))
     assert isinstance(result, str)
 
 
 def test_invoke() -> None:
     """Test invoke tokens from OllamaLLM."""
     llm = OllamaLLM(model=MODEL_NAME)
-
-    result = llm.invoke("I'm Pickle Rick", config=dict(tags=["foo"]))
+    result = llm.invoke("I'm Pickle Rick", config=RunnableConfig(tags=["foo"]))
     assert isinstance(result, str)

libs/partners/ollama/uv.lock

@@ -363,7 +363,7 @@ typing = [
 
 [[package]]
 name = "langchain-ollama"
-version = "0.3.3"
+version = "0.3.4"
 source = { editable = "." }
 dependencies = [
     { name = "langchain-core" },
@@ -397,7 +397,7 @@ typing = [
 [package.metadata]
 requires-dist = [
     { name = "langchain-core", editable = "../../core" },
-    { name = "ollama", specifier = ">=0.4.8,<1.0.0" },
+    { name = "ollama", specifier = ">=0.5.1,<1.0.0" },
 ]
 
 [package.metadata.requires-dev]
@@ -795,15 +795,15 @@ wheels = [
 
 [[package]]
 name = "ollama"
-version = "0.4.8"
+version = "0.5.1"
 source = { registry = "https://pypi.org/simple" }
 dependencies = [
     { name = "httpx" },
     { name = "pydantic" },
 ]
-sdist = { url = "https://files.pythonhosted.org/packages/e2/64/709dc99030f8f46ec552f0a7da73bbdcc2da58666abfec4742ccdb2e800e/ollama-0.4.8.tar.gz", hash = "sha256:1121439d49b96fa8339842965d0616eba5deb9f8c790786cdf4c0b3df4833802", size = 12972, upload-time = "2025-04-16T21:55:14.101Z" }
+sdist = { url = "https://files.pythonhosted.org/packages/8d/96/c7fe0d2d1b3053be614822a7b722c7465161b3672ce90df71515137580a0/ollama-0.5.1.tar.gz", hash = "sha256:5a799e4dc4e7af638b11e3ae588ab17623ee019e496caaf4323efbaa8feeff93", size = 41112, upload-time = "2025-05-30T21:32:48.679Z" }
 wheels = [
-    { url = "https://files.pythonhosted.org/packages/33/3f/164de150e983b3a16e8bf3d4355625e51a357e7b3b1deebe9cc1f7cb9af8/ollama-0.4.8-py3-none-any.whl", hash = "sha256:04312af2c5e72449aaebac4a2776f52ef010877c554103419d3f36066fe8af4c", size = 13325, upload-time = "2025-04-16T21:55:12.779Z" },
+    { url = "https://files.pythonhosted.org/packages/d6/76/3f96c8cdbf3955d7a73ee94ce3e0db0755d6de1e0098a70275940d1aff2f/ollama-0.5.1-py3-none-any.whl", hash = "sha256:4c8839f35bc173c7057b1eb2cbe7f498c1a7e134eafc9192824c8aecb3617506", size = 13369, upload-time = "2025-05-30T21:32:47.429Z" },
 ]
 
 [[package]]