feat(openai): custom tools (#32449)

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",

libs/core/langchain_core/tools/base.py

@@ -74,7 +74,14 @@ if TYPE_CHECKING:
     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):

libs/core/langchain_core/utils/function_calling.py

@@ -575,12 +575,23 @@ def convert_to_openai_tool(
 
         Added support for OpenAI's image generation built-in tool.
     """
+    from langchain_core.tools import Tool
+
     if isinstance(tool, dict):
         if tool.get("type") in _WellKnownOpenAITools:
             return tool
         # As of 03.12.25 can be "web_search_preview" or "web_search_preview_2025_03_11"
         if (tool.get("type") or "").startswith("web_search_preview"):
             return tool
+    if isinstance(tool, Tool) and (tool.metadata or {}).get("type") == "custom_tool":
+        oai_tool = {
+            "type": "custom",
+            "name": tool.name,
+            "description": tool.description,
+        }
+        if tool.metadata is not None and "format" in tool.metadata:
+            oai_tool["format"] = tool.metadata["format"]
+        return oai_tool
     oai_function = convert_to_openai_function(tool, strict=strict)
     return {"type": "function", "function": oai_function}
 

libs/partners/openai/langchain_openai/__init__.py

@@ -1,6 +1,7 @@
 from langchain_openai.chat_models import AzureChatOpenAI, ChatOpenAI
 from langchain_openai.embeddings import AzureOpenAIEmbeddings, OpenAIEmbeddings
 from langchain_openai.llms import AzureOpenAI, OpenAI
+from langchain_openai.tools import custom_tool
 
 __all__ = [
     "OpenAI",
@@ -9,4 +10,5 @@ __all__ = [
     "AzureOpenAI",
     "AzureChatOpenAI",
     "AzureOpenAIEmbeddings",
+    "custom_tool",
 ]

libs/partners/openai/langchain_openai/chat_models/base.py

@@ -3582,6 +3582,20 @@ def _make_computer_call_output_from_message(message: ToolMessage) -> dict:
     return computer_call_output
 
 
+def _make_custom_tool_output_from_message(message: ToolMessage) -> Optional[dict]:
+    custom_tool_output = None
+    for block in message.content:
+        if isinstance(block, dict) and block.get("type") == "custom_tool_call_output":
+            custom_tool_output = {
+                "type": "custom_tool_call_output",
+                "call_id": message.tool_call_id,
+                "output": block.get("output") or "",
+            }
+            break
+
+    return custom_tool_output
+
+
 def _pop_index_and_sub_index(block: dict) -> dict:
     """When streaming, langchain-core uses the ``index`` key to aggregate
     text blocks. OpenAI API does not support this key, so we need to remove it.
@@ -3608,7 +3622,10 @@ def _construct_responses_api_input(messages: Sequence[BaseMessage]) -> list:
             msg.pop("name")
         if msg["role"] == "tool":
             tool_output = msg["content"]
-            if lc_msg.additional_kwargs.get("type") == "computer_call_output":
+            custom_tool_output = _make_custom_tool_output_from_message(lc_msg)  # type: ignore[arg-type]
+            if custom_tool_output:
+                input_.append(custom_tool_output)
+            elif lc_msg.additional_kwargs.get("type") == "computer_call_output":
                 computer_call_output = _make_computer_call_output_from_message(
                     cast(ToolMessage, lc_msg)
                 )
@@ -3663,6 +3680,7 @@ def _construct_responses_api_input(messages: Sequence[BaseMessage]) -> list:
                             "file_search_call",
                             "function_call",
                             "computer_call",
+                            "custom_tool_call",
                             "code_interpreter_call",
                             "mcp_call",
                             "mcp_list_tools",
@@ -3690,7 +3708,8 @@ def _construct_responses_api_input(messages: Sequence[BaseMessage]) -> list:
                 content_call_ids = {
                     block["call_id"]
                     for block in input_
-                    if block.get("type") == "function_call" and "call_id" in block
+                    if block.get("type") in ("function_call", "custom_tool_call")
+                    and "call_id" in block
                 }
                 for tool_call in tool_calls:
                     if tool_call["id"] not in content_call_ids:
@@ -3841,6 +3860,15 @@ def _construct_lc_result_from_responses_api(
                     "error": error,
                 }
                 invalid_tool_calls.append(tool_call)
+        elif output.type == "custom_tool_call":
+            content_blocks.append(output.model_dump(exclude_none=True, mode="json"))
+            tool_call = {
+                "type": "tool_call",
+                "name": output.name,
+                "args": {"__arg1": output.input},
+                "id": output.call_id,
+            }
+            tool_calls.append(tool_call)
         elif output.type in (
             "reasoning",
             "web_search_call",
@@ -4044,6 +4072,23 @@ def _convert_responses_chunk_to_generation_chunk(
         tool_output = chunk.item.model_dump(exclude_none=True, mode="json")
         tool_output["index"] = current_index
         content.append(tool_output)
+    elif (
+        chunk.type == "response.output_item.done"
+        and chunk.item.type == "custom_tool_call"
+    ):
+        _advance(chunk.output_index)
+        tool_output = chunk.item.model_dump(exclude_none=True, mode="json")
+        tool_output["index"] = current_index
+        content.append(tool_output)
+        tool_call_chunks.append(
+            {
+                "type": "tool_call_chunk",
+                "name": chunk.item.name,
+                "args": json.dumps({"__arg1": chunk.item.input}),
+                "id": chunk.item.call_id,
+                "index": current_index,
+            }
+        )
     elif chunk.type == "response.function_call_arguments.delta":
         _advance(chunk.output_index)
         tool_call_chunks.append(

libs/partners/openai/langchain_openai/tools/__init__.py

@@ -0,0 +1,3 @@
+from langchain_openai.tools.custom_tool import custom_tool
+
+__all__ = ["custom_tool"]

libs/partners/openai/langchain_openai/tools/custom_tool.py

@@ -0,0 +1,109 @@
+import inspect
+from collections.abc import Awaitable
+from typing import Any, Callable
+
+from langchain_core.tools import tool
+
+
+def _make_wrapped_func(func: Callable[..., str]) -> Callable[..., list[dict[str, Any]]]:
+    def wrapped(x: str) -> list[dict[str, Any]]:
+        return [{"type": "custom_tool_call_output", "output": func(x)}]
+
+    return wrapped
+
+
+def _make_wrapped_coroutine(
+    coroutine: Callable[..., Awaitable[str]],
+) -> Callable[..., Awaitable[list[dict[str, Any]]]]:
+    async def wrapped(*args: Any, **kwargs: Any) -> list[dict[str, Any]]:
+        result = await coroutine(*args, **kwargs)
+        return [{"type": "custom_tool_call_output", "output": result}]
+
+    return wrapped
+
+
+def custom_tool(*args: Any, **kwargs: Any) -> Any:
+    """Decorator to create an OpenAI custom tool.
+
+    Custom tools allow for tools with (potentially long) freeform string inputs.
+
+    See below for an example using LangGraph:
+
+    .. code-block:: python
+
+        @custom_tool
+        def execute_code(code: str) -> str:
+            \"\"\"Execute python code.\"\"\"
+            return "27"
+
+
+        llm = ChatOpenAI(model="gpt-5", output_version="responses/v1")
+
+        agent = create_react_agent(llm, [execute_code])
+
+        input_message = {"role": "user", "content": "Use the tool to calculate 3^3."}
+        for step in agent.stream(
+            {"messages": [input_message]},
+            stream_mode="values",
+        ):
+            step["messages"][-1].pretty_print()
+
+    You can also specify a format for a corresponding context-free grammar using the
+    ``format`` kwarg:
+
+    .. code-block:: python
+
+        from langchain_openai import ChatOpenAI, custom_tool
+        from langgraph.prebuilt import create_react_agent
+
+        grammar = \"\"\"
+        start: expr
+        expr: term (SP ADD SP term)* -> add
+        | term
+        term: factor (SP MUL SP factor)* -> mul
+        | factor
+        factor: INT
+        SP: " "
+        ADD: "+"
+        MUL: "*"
+        %import common.INT
+        \"\"\"
+
+        format = {"type": "grammar", "syntax": "lark", "definition": grammar}
+
+        # highlight-next-line
+        @custom_tool(format=format)
+        def do_math(input_string: str) -> str:
+            \"\"\"Do a mathematical operation.\"\"\"
+            return "27"
+
+
+        llm = ChatOpenAI(model="gpt-5", output_version="responses/v1")
+
+        agent = create_react_agent(llm, [do_math])
+
+        input_message = {"role": "user", "content": "Use the tool to calculate 3^3."}
+        for step in agent.stream(
+            {"messages": [input_message]},
+            stream_mode="values",
+        ):
+            step["messages"][-1].pretty_print()
+    """
+
+    def decorator(func: Callable[..., Any]) -> Any:
+        metadata = {"type": "custom_tool"}
+        if "format" in kwargs:
+            metadata["format"] = kwargs.pop("format")
+        tool_obj = tool(infer_schema=False, **kwargs)(func)
+        tool_obj.metadata = metadata
+        tool_obj.description = func.__doc__
+        if inspect.iscoroutinefunction(func):
+            tool_obj.coroutine = _make_wrapped_coroutine(func)
+        else:
+            tool_obj.func = _make_wrapped_func(func)
+        return tool_obj
+
+    if args and callable(args[0]) and not kwargs:
+        return decorator(args[0])
+
+    return decorator

libs/partners/openai/tests/integration_tests/chat_models/test_responses_api.py

@@ -17,7 +17,7 @@ from langchain_core.messages import (
 from pydantic import BaseModel
 from typing_extensions import TypedDict
 
-from langchain_openai import ChatOpenAI
+from langchain_openai import ChatOpenAI, custom_tool
 
 MODEL_NAME = "gpt-4o-mini"
 
@@ -672,3 +672,32 @@ def test_image_generation_multi_turn() -> None:
     _check_response(ai_message2)
     tool_output2 = ai_message2.additional_kwargs["tool_outputs"][0]
     assert set(tool_output2.keys()).issubset(expected_keys)
+
+
+@pytest.mark.vcr()
+def test_custom_tool() -> None:
+    @custom_tool
+    def execute_code(code: str) -> str:
+        """Execute python code."""
+        return "27"
+
+    llm = ChatOpenAI(model="gpt-5", output_version="responses/v1").bind_tools(
+        [execute_code]
+    )
+
+    input_message = {"role": "user", "content": "Use the tool to evaluate 3^3."}
+    tool_call_message = llm.invoke([input_message])
+    assert isinstance(tool_call_message, AIMessage)
+    assert len(tool_call_message.tool_calls) == 1
+    tool_call = tool_call_message.tool_calls[0]
+    tool_message = execute_code.invoke(tool_call)
+    response = llm.invoke([input_message, tool_call_message, tool_message])
+    assert isinstance(response, AIMessage)
+
+    # Test streaming
+    full: Optional[BaseMessageChunk] = None
+    for chunk in llm.stream([input_message]):
+        assert isinstance(chunk, AIMessageChunk)
+        full = chunk if full is None else full + chunk
+    assert isinstance(full, AIMessageChunk)
+    assert len(full.tool_calls) == 1

libs/partners/openai/tests/unit_tests/test_imports.py

@@ -7,6 +7,7 @@ EXPECTED_ALL = [
     "AzureOpenAI",
     "AzureChatOpenAI",
     "AzureOpenAIEmbeddings",
+    "custom_tool",
 ]
 
 

libs/partners/openai/tests/unit_tests/test_tools.py

@@ -0,0 +1,120 @@
+from langchain_core.messages import AIMessage, HumanMessage, ToolMessage
+from langchain_core.tools import Tool
+
+from langchain_openai import ChatOpenAI, custom_tool
+
+
+def test_custom_tool() -> None:
+    @custom_tool
+    def my_tool(x: str) -> str:
+        """Do thing."""
+        return "a" + x
+
+    # Test decorator
+    assert isinstance(my_tool, Tool)
+    assert my_tool.metadata == {"type": "custom_tool"}
+    assert my_tool.description == "Do thing."
+
+    result = my_tool.invoke(
+        {
+            "type": "tool_call",
+            "name": "my_tool",
+            "args": {"whatever": "b"},
+            "id": "abc",
+            "extras": {"type": "custom_tool_call"},
+        }
+    )
+    assert result == ToolMessage(
+        [{"type": "custom_tool_call_output", "output": "ab"}],
+        name="my_tool",
+        tool_call_id="abc",
+    )
+
+    # Test tool schema
+    ## Test with format
+    @custom_tool(format={"type": "grammar", "syntax": "lark", "definition": "..."})
+    def another_tool(x: str) -> None:
+        """Do thing."""
+        pass
+
+    llm = ChatOpenAI(model="gpt-4.1", use_responses_api=True).bind_tools([another_tool])
+    assert llm.kwargs == {  # type: ignore[attr-defined]
+        "tools": [
+            {
+                "type": "custom",
+                "name": "another_tool",
+                "description": "Do thing.",
+                "format": {"type": "grammar", "syntax": "lark", "definition": "..."},
+            }
+        ]
+    }
+
+    llm = ChatOpenAI(model="gpt-4.1", use_responses_api=True).bind_tools([my_tool])
+    assert llm.kwargs == {  # type: ignore[attr-defined]
+        "tools": [{"type": "custom", "name": "my_tool", "description": "Do thing."}]
+    }
+
+    # Test passing messages back
+    message_history = [
+        HumanMessage("Use the tool"),
+        AIMessage(
+            [
+                {
+                    "type": "custom_tool_call",
+                    "id": "ctc_abc123",
+                    "call_id": "abc",
+                    "name": "my_tool",
+                    "input": "a",
+                }
+            ],
+            tool_calls=[
+                {
+                    "type": "tool_call",
+                    "name": "my_tool",
+                    "args": {"__arg1": "a"},
+                    "id": "abc",
+                }
+            ],
+        ),
+        result,
+    ]
+    payload = llm._get_request_payload(message_history)  # type: ignore[attr-defined]
+    expected_input = [
+        {"content": "Use the tool", "role": "user"},
+        {
+            "type": "custom_tool_call",
+            "id": "ctc_abc123",
+            "call_id": "abc",
+            "name": "my_tool",
+            "input": "a",
+        },
+        {"type": "custom_tool_call_output", "call_id": "abc", "output": "ab"},
+    ]
+    assert payload["input"] == expected_input
+
+
+async def test_async_custom_tool() -> None:
+    @custom_tool
+    async def my_async_tool(x: str) -> str:
+        """Do async thing."""
+        return "a" + x
+
+    # Test decorator
+    assert isinstance(my_async_tool, Tool)
+    assert my_async_tool.metadata == {"type": "custom_tool"}
+    assert my_async_tool.description == "Do async thing."
+
+    result = await my_async_tool.ainvoke(
+        {
+            "type": "tool_call",
+            "name": "my_async_tool",
+            "args": {"whatever": "b"},
+            "id": "abc",
+            "extras": {"type": "custom_tool_call"},
+        }
+    )
+    assert result == ToolMessage(
+        [{"type": "custom_tool_call_output", "output": "ab"}],
+        name="my_async_tool",
+        tool_call_id="abc",
+    )