{
"$type": "site.standard.document",
"bskyPostRef": {
"cid": "bafyreic7erfvspjcb5zjrba2gwyaewi2asz7ad3v7br6r4ckqzlyx64tyu",
"uri": "at://did:plc:lk3jfj3zq4k4wxnk474axylu/app.bsky.feed.post/3mmcwcg3wgxs2"
},
"path": "/t/open-ai-realtime-ws-connections-are-cut-off-continuously/1381378#post_3",
"publishedAt": "2026-05-20T20:45:23.000Z",
"site": "https://community.openai.com",
"textContent": "Yes, I can reproduce this.\n\nWith `OpenAI-Beta: realtime=v1`, the WebSocket upgrade succeeds, but OpenAI then sends an error event and closes the connection with:\n\n\n code=4000\n reason=invalid_request_error.beta_api_shape_disabled\n\n\nUsing the same endpoint and model without the beta header succeeds and returns `session.created`.\n\nSo the practical next step appears to be removing:\n\n\n OpenAI-Beta: realtime=v1\n\n\nAnd then updating any beta Realtime event or session shapes to the GA version.",
"title": "Open AI realtime WS connections are cut off continuously"
}