{
  "schema_version": "1.0",
  "id": "s4:https://vercel.com/changelog/websocket-support-for-openai-responses-api-live-on-ai-gateway",
  "slug": "websocket-support-for-openai-responses-api-live-on-ai-gateway-0oj5npd",
  "url": "https://feed7.dev/p/websocket-support-for-openai-responses-api-live-on-ai-gateway-0oj5npd",
  "title": "WebSocket support for OpenAI Responses API live on AI Gateway",
  "why_included": "Vercel’s gateway now supports persistent WebSocket sessions for the Responses API, reducing repeated context transfer during long, tool-heavy agent runs.",
  "summary": "AI Gateway now exposes the OpenAI Responses API over **WebSocket** at GET /v1/responses. Each new turn can send fresh input plus **previous_response_id** instead of resubmitting the full context over HTTP.",
  "practical_implication": "Consider the persistent route for long coding-agent loops where connection and context overhead accumulate. OpenAI reports **up to about 40% faster** end-to-end execution for rollouts with at least 20 tool calls.",
  "agent_context": "AI Gateway now exposes the OpenAI Responses API over **WebSocket** at GET /v1/responses. Each new turn can send fresh input plus **previous_response_id** instead of resubmitting the full context over HTTP.\n\nConsider the persistent route for long coding-agent loops where connection and context overhead accumulate. OpenAI reports **up to about 40% faster** end-to-end execution for rollouts with at least 20 tool calls.\n\nThat performance figure applies to a specific tool-heavy workload, not every interaction. The route accepts raw response.create frames and supports **store=false and Zero Data Retention**, so clients must manage the socket and event protocol.",
  "source": {
    "name": "Vercel",
    "url": "https://vercel.com/changelog/websocket-support-for-openai-responses-api-live-on-ai-gateway",
    "published_at": "2026-07-27T00:01:00.000Z"
  },
  "source_class": "blog_post",
  "content_type": "Engineering Post",
  "layer": "infra",
  "domains": [
    "coding"
  ],
  "topics": [
    "gateways",
    "agent-reliability"
  ],
  "verification": {
    "status": "official_source",
    "label": "Official Source",
    "method": "source_feed",
    "verified_at": null
  },
  "uncertainty": [
    "That performance figure applies to a specific tool-heavy workload, not every interaction. The route accepts raw response.create frames and supports **store=false and Zero Data Retention**, so clients must manage the socket and event protocol."
  ],
  "lifecycle": "Current",
  "published_at": "2026-07-27T00:01:00.000Z",
  "modified_at": "2026-07-27T00:01:00.000Z",
  "supersedes": [],
  "expires_at": null,
  "formats": {
    "html": "https://feed7.dev/p/websocket-support-for-openai-responses-api-live-on-ai-gateway-0oj5npd",
    "json": "https://feed7.dev/p/websocket-support-for-openai-responses-api-live-on-ai-gateway-0oj5npd.json",
    "markdown": "https://feed7.dev/p/websocket-support-for-openai-responses-api-live-on-ai-gateway-0oj5npd.md"
  }
}