{
"$type": "site.standard.document",
"bskyPostRef": {
"cid": "bafyreig6f3uvfdcdrx6vebinkxixwqdy5y3pcrdvhvdymdehtojinr2yx4",
"uri": "at://did:plc:lk3jfj3zq4k4wxnk474axylu/app.bsky.feed.post/3mg6vd7rgoyy2"
},
"path": "/t/error-when-attempting-to-access-gpt-image-1-5-in-images-edits/1375534#post_14",
"publishedAt": "2026-03-03T21:08:18.000Z",
"site": "https://community.openai.com",
"textContent": "The docs for “edits” simply “never said” what was needed at all - the prior set of parameters is still around, and it is easy to make a mistaken assumption that if you want ‘b64’ you still might ask for it. Missing:\n\n_j:\n\n> Unsupported for `gpt-image-1`.\n\nThen, if you are forced to infer parameter usage from the “generations” docs, you might also wonder why “quality” of gpt-image models doesn’t work on the edits API, either.\n\nHere’s the “edits” API documentation with “response_format” that was wiped, instead of properly modified with notations.\n\n\n model:\n anyOf:\n - type: string\n - type: string\n enum:\n - dall-e-2\n x-stainless-const: true\n x-oaiTypeLabel: string\n nullable: true\n description: The model to use for image generation. Only `dall-e-2` is supported at this time.\n 'n':\n type: integer\n minimum: 1\n maximum: 10\n default: 1\n example: 1\n nullable: true\n description: The number of images to generate. Must be between 1 and 10.\n response_format:\n type: string\n enum:\n - url\n - b64_json\n default: url\n example: url\n nullable: true\n description: >-\n The format in which the generated images are returned. Must be one of `url` or `b64_json`. URLs\n are only valid for 60 minutes after the image has been generated.\n",
"title": "Error when attempting to access gpt-image-1.5 in images/edits"
}