Streaming Guide

Design resilient clients for SSE streaming across model protocols.

Streaming

For OpenAI-compatible and Claude-compatible requests, set stream: true in the JSON body. Gemini streaming can use :streamGenerateContent, alt=sse, or model/client-specific streaming helpers. Streaming responses are delivered as SSE chunks; handle network disconnects and mid-stream errors explicitly.

curl https://onerouter.com/v1/chat/completions \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "openai/gpt-5.2",
    "messages": [{"role": "user", "content": "Write a short launch note."}],
    "stream": true
  }'

Consume a stream

With an OpenAI-compatible SDK, iterate the streamed chunks and read choices[0].delta.content as it arrives. Print or forward each fragment to build the response incrementally.

from openai import OpenAI

client = OpenAI(base_url="https://onerouter.com/v1", api_key="YOUR_API_KEY")

stream = client.chat.completions.create(
    model="openai/gpt-5.2",
    messages=[{"role": "user", "content": "Write a short launch note."}],
    stream=True,
)

for chunk in stream:
    delta = chunk.choices[0].delta.content or ""
    print(delta, end="", flush=True)

Client behavior

Treat streaming as a long-lived network operation. Parse SSE frames incrementally, ignore keep-alive comments, surface mid-stream errors to users, and let users cancel the request. If a stream disconnects before a final completion event, check request logs and retry only if your operation is safe to repeat.