{
"$type": "site.standard.document",
"bskyPostRef": {
"cid": "bafyreibihtjsf4gvztsx3auw2f7u6ucznqrxhlyxuj7w2m23qwrkprxxvq",
"uri": "at://did:plc:lk3jfj3zq4k4wxnk474axylu/app.bsky.feed.post/3meoyzbimrvw2"
},
"path": "/t/gpt-5-3-codex-spark-research-preview-with-1000-tokens-per-second/1374091#post_3",
"publishedAt": "2026-02-12T20:56:45.000Z",
"site": "https://community.openai.com",
"tags": [
"x.com",
"Steve ☕️",
"@stevendcoffey",
"@brianyu8",
"@ashwinnathan09",
"pic.x.com/cVH1eOobYw",
"6:14 PM - 12 Feb 2026"
],
"textContent": "Apparently, while working on Spark, something was implemented that many more developers will benefit from:\n\nx.com\n\n#### Steve ☕️\n\n@stevendcoffey\n\nHard to overstate just how much work went into this one sentence. HUGE props to @brianyu8, @ashwinnathan09 and team on making these huge improvements Websockets in /v1/responses landing in GA soon! pic.x.com/cVH1eOobYw\n\n6:14 PM - 12 Feb 2026 46 2",
"title": "GPT- 5.3-Codex-Spark Research Preview with 1000 Tokens per Second"
}