{
  "$type": "site.standard.document",
  "bskyPostRef": {
    "cid": "bafyreibga23653ttqrvp2wzsif5gdaje2jq76ybcxf5zqx6bot6mro2diq",
    "uri": "at://did:plc:lk3jfj3zq4k4wxnk474axylu/app.bsky.feed.post/3mlvvot5tuoi2"
  },
  "path": "/t/why-do-gpt-5-1-and-gpt-5-4-mini-behave-so-differently-in-production-chatbot-use-cases/1380891#post_5",
  "publishedAt": "2026-05-15T17:13:35.000Z",
  "site": "https://community.openai.com",
  "tags": [
    "@vb"
  ],
  "textContent": "@vb You touched on a very interesting point. Actually, I haven’t tested changing `reasoning.effort` on either model yet. I thought about it, but I didn’t try it in practice.\n\nFrom what I found in the OpenAI docs, `gpt-5.1` supports `none`, `low`, `medium`, and `high`, and its default is `none`. For `gpt-5.4-mini`, the model page shows support for `none`, `low`, `medium`, `high`, and `xhigh`, but I didn’t find an explicit default stated there.\n\nDo you recommend testing `gpt-5.4-mini` with a higher `reasoning.effort`, such as `high` or `xhigh`?\n\nMy main concern is: if I increase `reasoning.effort` on `gpt-5.4-mini`, could it end up with similar latency/cost conditions as `gpt-5.1`, reducing the advantage of using the mini model?",
  "title": "Why do gpt-5.1 and gpt-5.4-mini behave so differently in production chatbot use cases?"
}