> ## Documentation Index
> Fetch the complete documentation index at: https://docs.meshai.dev/llms.txt
> Use this file to discover all available pages before exploring further.

# LangChain / LangGraph

> Auto-track LLM calls in LangChain chains and LangGraph agents

## Installation

```bash theme={null}
pip install meshai-sdk[langchain]
```

## Usage with LangChain

```python theme={null}
from meshai import MeshAI
from meshai.integrations.langchain import MeshAICallbackHandler
from langchain_openai import ChatOpenAI

client = MeshAI(api_key="msh_...", agent_name="my-chain")
client.register(framework="langchain")

handler = MeshAICallbackHandler(client)

# Pass to any LangChain model
llm = ChatOpenAI(model="gpt-4o", callbacks=[handler])
response = llm.invoke("What is MeshAI?")
# Model and tokens tracked automatically
```

## Usage with LangGraph

```python theme={null}
from langgraph.graph import StateGraph
from meshai.integrations.langchain import MeshAICallbackHandler

handler = MeshAICallbackHandler(client)

# Pass as config to graph execution
config = {"callbacks": [handler]}
result = graph.stream(input_data, config=config)
```

## Usage with Chains

```python theme={null}
from langchain_core.prompts import ChatPromptTemplate

prompt = ChatPromptTemplate.from_messages([...])
chain = prompt | llm  # llm already has the handler from callbacks=[handler]
result = chain.invoke({"topic": "AI governance"})
```

## How It Works

`MeshAICallbackHandler` implements LangChain's callback interface. The `on_llm_end` method:

1. Extracts the **model name** from `LLMResult.llm_output`
2. Extracts **token usage** from the response metadata
3. Infers the **provider** from the model name
4. Sends to MeshAI (buffered, non-blocking)

Works with any LangChain-compatible model: ChatOpenAI, ChatAnthropic, ChatGoogle, etc.