Checked other resources

• This is a bug, not a usage question.
• I added a clear and descriptive title that summarizes this issue.
• I used the GitHub search to find a similar question and didn't find it.
• I am sure that this is a bug in LangChain rather than my code.
• The bug is not resolved by updating to the latest stable version of LangChain (or the specific integration package).
• This is not related to the langchain-community package.
• I posted a self-contained, minimal, reproducible example. A maintainer can copy it and run it AS IS.

Package (Required)

• langchain
• langchain-openai
• langchain-anthropic
• langchain-classic
• langchain-core
• langchain-model-profiles
• langchain-tests
• langchain-text-splitters
• langchain-chroma
• langchain-deepseek
• langchain-exa
• langchain-fireworks
• langchain-groq
• langchain-huggingface
• langchain-mistralai
• langchain-nomic
• langchain-ollama
• langchain-perplexity
• langchain-prompty
• langchain-qdrant
• langchain-xai
• Other / not sure / general

Reproduction Steps / Example Code (Python)

from langchain_core.output_parsers.openai_tools import PydanticToolsParser
from pydantic import BaseModel
from langchain_core.messages import AIMessage
from langchain_core.outputs import ChatGeneration

class Model1(BaseModel):
    name: str
    age: int

class Model2(BaseModel):
    value: float

parser = PydanticToolsParser(tools=[Model1, Model2])
result = ChatGeneration(
    message=AIMessage(
        content="",
        additional_kwargs={
            "tool_calls": [
                {"type": "function", "function": {"name": "Model1", "arguments": '{"name": "Alice", "age": 30}'}},
                {"type": "function", "function": {"name": "Model3", "arguments": '{"value": 3.14}'}}
            ]
        }
    )
)
output = parser.parse_result([result], partial=True)

Error Message and Stack Trace (if applicable)

Traceback (most recent call last):
  File "/home/workplace/lc4.py", line 25, in <module>
    output = parser.parse_result([result], partial=True)
  File "/lib/python3.10/site-packages/langchain_core/output_parsers/openai_tools.py", line 306, in parse_result
    pydantic_objects.append(name_dict[res["type"]](**res["args"]))
KeyError: 'Model3'

Description

When using PydanticToolsParser.parse_result with partial=True, if the LLM output contains a tool call referencing a tool that is not registered in the parser, a KeyError is raised.

From the API documentation, partial=True is expected to allow safely skipping partially parseable outputs. However, the current implementation does not account for unmatched tool names, so even with partial=True, the method crashes instead of skipping the unknown tool.

This seems to be purely an internal logic oversight: the code checks partial for argument type or validation errors, but it does not handle the case where res["type"] is not in name_dict, leading to an unhandled KeyError.

System Info

Package Information

langchain_core: 0.3.68
langchain: 0.3.26
langsmith: 0.4.5
langchain_text_splitters: 0.3.8