ドキュメント · 導入

ストリーミング

Set stream: true to receive tokens as server-sent events (SSE) while they are generated. You get the first token sooner and avoid client timeouts on long outputs.

本文は現在英語版のみ提供しています。

curl https://<your-endpoint>/v1/chat/completions \
  -H "Authorization: Bearer $API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-4.1-mini",
    "messages": [{"role": "user", "content": "Hello!"}],
    "stream": true
  }'

Event format

Each event is a data: {JSON} line with incremental text in choices[0].delta.content. The stream ends with data: [DONE].

data: {"id":"chatcmpl-...","choices":[{"index":0,"delta":{"role":"assistant","content":""}}]}

data: {"id":"chatcmpl-...","choices":[{"index":0,"delta":{"content":"Hel"}}]}

data: {"id":"chatcmpl-...","choices":[{"index":0,"delta":{"content":"lo!"},"finish_reason":"stop"}]}

data: [DONE]

Usage in streams

Set stream_options.include_usage and the final chunk carries the usage for the whole request.

stream = client.chat.completions.create(
    model="gpt-4.1-mini",
    messages=[{"role": "user", "content": "Write a haiku about routers."}],
    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:            # final chunk: empty choices, carries usage
        print("\n", chunk.usage)

Things to know

If you relay streams through your own reverse proxy (e.g. Nginx), disable response buffering (proxy_buffering off), or the content arrives all at once at the end.
Set the read timeout for streams as the gap between chunks, not the total request duration.
Reasoning models may think for a while before the first token. That is expected.
In the Anthropic format, stream events match Anthropic's (message_start, content_block_delta, …) — see Anthropic format.

チャット補完

Anthropic 形式