# Streaming

> Receive tokens as they are generated with server-sent events.

Set `stream: true` to receive tokens as server-sent events (SSE) instead of waiting for the full response.

```python
stream = client.chat.completions.create(
    model="deepseek-ai/DeepSeek-V4.1-Flash",
    messages=[{"role": "user", "content": "Write a haiku about rivers."}],
    stream=True,
    stream_options={"include_usage": True},
)
for chunk in stream:
    if chunk.choices and chunk.choices[0].delta.content:
        print(chunk.choices[0].delta.content, end="", flush=True)
    if chunk.usage:
        print("\n", chunk.usage)
```

## What to expect

- Each event is a `data:` line containing a chat completion chunk; the stream ends with `data: [DONE]`.
- With `stream_options.include_usage`, the final chunk carries the `usage` object. Kurrens always reports usage for
  streamed requests.
- Reasoning models may send SSE comment lines (`: keep-alive`) while they think. Clients should ignore them — they keep
  the connection from timing out.
- If an error occurs mid-stream, a final event contains an `error` object before the stream closes.

## Cancelling

Closing the connection cancels generation. You're billed only for tokens generated before cancellation.