{
  "$type": "site.standard.document",
  "bskyPostRef": {
    "cid": "bafyreiateejilowqnj2lyuhkjjhkyvaadmrpvkysdmhhymopa5mdddtcou",
    "uri": "at://did:plc:lk3jfj3zq4k4wxnk474axylu/app.bsky.feed.post/3meqvhvtxm4g2"
  },
  "path": "/t/gpt-5-3-codex-spark-research-preview-with-1000-tokens-per-second/1374091#post_4",
  "publishedAt": "2026-02-13T15:32:56.000Z",
  "site": "https://community.openai.com",
  "textContent": "This looks fantastic but having started several projects with the base codex, you are warned not to switch to this model.\n\nI look forward to my next inspiration and I will give this a go from the get-go!\n\nReal-time?\n\nKeen to hear people’s impressions esp. when compared to codex medium and Claude.",
  "title": "GPT- 5.3-Codex-Spark Research Preview with 1000 Tokens per Second"
}