{
  "$type": "site.standard.document",
  "bskyPostRef": {
    "cid": "bafyreigpfebwr4qy3l5nmu4xg7ra7nxkxcp43smqypwvf27ltwcbm63q5a",
    "uri": "at://did:plc:lk3jfj3zq4k4wxnk474axylu/app.bsky.feed.post/3mhl74cykaig2"
  },
  "path": "/t/the-new-speed-feature-for-codex-what-is-your-experience/1377408#post_2",
  "publishedAt": "2026-03-21T12:32:51.000Z",
  "site": "https://community.openai.com",
  "tags": [
    "GPT- 5.3-Codex-Spark Research Preview with 1000 Tokens per Second",
    "Codex",
    "ultra-fast model for real-time coding",
    "[SWE-Bench Pro-1]"
  ],
  "textContent": "Apologies for the sidestep, but if you want something really fast and can handle less thinking and slightly less accuracy, try:\n\nGPT- 5.3-Codex-Spark Research Preview with 1000 Tokens per Second Codex\n\n> An ultra-fast model for real-time coding rolling out today to ChatGPT Pro users in the Codex app, CLI, and IDE extension. From the announcement: Codex-Spark marks the first milestone in our partnership with Cerebras, which we announced in January⁠. Codex-Spark is optimized to feel near-instant when served on ultra-low latency hardware—delivering more than 1000 tokens per second while remaining highly capable for real-world coding tasks. [SWE-Bench Pro-1] At launch, Codex-Spark has a 128k…\n\n(if you have access)\n\nBecause it is very _very_ quick.",
  "title": "The new speed feature for Codex . What is your experience?"
}