> Fetch clean Markdown by appending `.md` to any page URL under https://signalwire.com/docs or requesting it with the HTTP header `Accept: text/markdown`. The root index at https://signalwire.com/docs/llms.txt lists the available documentation indexes.

# create_conversation

> Create a conversation against your agent's SWML, or reset an existing one.

[ref-aichatclient]: /docs/server-sdks/reference/python/agents/ai-chat-client

[chat]: /docs/server-sdks/reference/python/agents/ai-chat-client/chat

Create a conversation, or reset an existing one when `reinit` is set. Calling it for an id that
already exists returns that conversation untouched.

## **Parameters**

**`conversation_id`** `str` — required

The conversation id. You choose it, and it is scoped to your project.

---

**`config_url`** `str` — required

The URL serving your agent's SWML. SignalWire fetches it server-to-server, so it must be reachable
from the public internet.

---

**`user_message`** `str | None` — default: None

An opening message from the user, sent as part of creating the conversation.

---

**`timeout`** `int | None` — default: None

Idle seconds before the conversation ends. Sent as `conversation_timeout`; the service default is
3600\.

---

**`user_metadata`** `dict[str, Any] | None` — default: None

Arbitrary data about the user, echoed back on this conversation's webhooks. Sent as
`user_meta_data`.

---

**`reinit`** `bool` — default: False

Reset an existing conversation instead of returning it as-is.

---

## **Returns**

`ConversationInfo` — carries `id`, `status` (`created`, `reinitialized`, or `exists`), and
`initial_message`.

> **Warning**
>
> `initial_message` is the agent's generated greeting. Create the conversation when the visitor
> opens the chat rather than when the page loads. Use [`chat()`][chat] with `config_url` instead when
> you don't need the agent to speak first.

## **Example**

```python {10}
import asyncio

from signalwire.ai_chat import AIChatClient

CONFIG_URL = "https://bayview-taxi.example.com/swml"


async def main():
    async with AIChatClient(space="your-space") as client:
        info = await client.create_conversation(
            "chat-8f21",
            config_url=CONFIG_URL,
            timeout=1800,
            user_metadata={"customer_tier": "premium"},
        )
        print(info.status, info.initial_message)


asyncio.run(main())
```

> Create a conversation against your agent's SWML, or reset an existing one.