# ResponsesHostServer

> **Class** in `langchain_azure_ai`

📖 [View in docs](https://reference.langchain.com/python/langchain-azure-ai/agents/hosting/_responses_host/ResponsesHostServer)

Host a LangGraph ``CompiledStateGraph`` as the Azure AI Responses API.

## Signature

```python
ResponsesHostServer(
    self,
    graph: 'CompiledStateGraph',
    *,
    app: Optional[ResponsesAgentServerHost] = None,
    options: Optional[ResponsesServerOptions] = None,
    store: Optional[ResponseProviderProtocol] = None,
    conversation_chain_store: Optional[ConversationChainStoreProtocol] = None,
    prefix: str = '',
    applicationinsights_connection_string: Optional[str] = None,
    graceful_shutdown_timeout: Optional[int] = None,
)
```

## Description

**Example:**

Create a LangChain agent graph and host it on ``POST /responses``::

    import os

    from langchain.agents import create_agent
    from langchain_openai import ChatOpenAI

    from langchain_azure_ai.agents.hosting import ResponsesHostServer

    model = ChatOpenAI(
        model=os.environ.get("AZURE_AI_MODEL_DEPLOYMENT_NAME", "gpt-4o"),
    )
    graph = create_agent(model, tools=[])

    ResponsesHostServer(graph).run(port=8088)

A minimal non-streaming request is ``{"input": "Hello!"}``.
Streaming requests use ``{"input": "Hello!", "stream": true}``.

The host owns an internal :class:`ResponsesAgentServerHost` and
registers a default request → graph → events conversion pipeline
against it. For advanced scenarios (custom routes, multi-protocol
composition, custom converter), users may either:

- subclass and override :meth:`handle_create`, or
- drop down to :class:`ResponsesAgentServerHost` directly and write
  their own ``@response_handler``.

## Parameters

| Name | Type | Required | Description |
|------|------|----------|-------------|
| `graph` | `'CompiledStateGraph'` | Yes | The compiled LangGraph state graph to host. By default the state schema must declare a ``messages`` field. Non- ``MessagesState`` graphs require subclassing and overriding :meth:`build_input` / :meth:`handle_create` (typically by reusing this class as a starting point). |

## Constructors

```python
__init__(
    self,
    graph: 'CompiledStateGraph',
    *,
    app: Optional[ResponsesAgentServerHost] = None,
    options: Optional[ResponsesServerOptions] = None,
    store: Optional[ResponseProviderProtocol] = None,
    conversation_chain_store: Optional[ConversationChainStoreProtocol] = None,
    prefix: str = '',
    applicationinsights_connection_string: Optional[str] = None,
    graceful_shutdown_timeout: Optional[int] = None,
) -> None
```

| Name | Type |
|------|------|
| `graph` | `'CompiledStateGraph'` |
| `app` | `Optional[ResponsesAgentServerHost]` |
| `options` | `Optional[ResponsesServerOptions]` |
| `store` | `Optional[ResponseProviderProtocol]` |
| `conversation_chain_store` | `Optional[ConversationChainStoreProtocol]` |
| `prefix` | `str` |
| `applicationinsights_connection_string` | `Optional[str]` |
| `graceful_shutdown_timeout` | `Optional[int]` |


## Properties

- `app`
- `graph`

## Methods

- [`run()`](https://reference.langchain.com/python/langchain-azure-ai/agents/hosting/_responses_host/ResponsesHostServer/run)
- [`run_async()`](https://reference.langchain.com/python/langchain-azure-ai/agents/hosting/_responses_host/ResponsesHostServer/run_async)
- [`build_input()`](https://reference.langchain.com/python/langchain-azure-ai/agents/hosting/_responses_host/ResponsesHostServer/build_input)
- [`build_resume_command()`](https://reference.langchain.com/python/langchain-azure-ai/agents/hosting/_responses_host/ResponsesHostServer/build_resume_command)
- [`detect_rejection()`](https://reference.langchain.com/python/langchain-azure-ai/agents/hosting/_responses_host/ResponsesHostServer/detect_rejection)
- [`build_runnable_config()`](https://reference.langchain.com/python/langchain-azure-ai/agents/hosting/_responses_host/ResponsesHostServer/build_runnable_config)
- [`build_runnable_config_sync()`](https://reference.langchain.com/python/langchain-azure-ai/agents/hosting/_responses_host/ResponsesHostServer/build_runnable_config_sync)
- [`handle_create()`](https://reference.langchain.com/python/langchain-azure-ai/agents/hosting/_responses_host/ResponsesHostServer/handle_create)

---

[View source on GitHub](https://github.com/langchain-ai/langchain-azure/blob/7a6a664d617264df8dcb12fcfb1bacc09f21d504/libs/azure-ai/langchain_azure_ai/agents/hosting/_responses_host.py#L248)