GPT-Image · Image
GPT-Image · Introduction
GPT-Image image generation, text-to-image and image-to-image. model is now required: gpt-image-1 / gpt-image-1.5 (4 credits/img), gpt-image-2 (6 credits/img). For size, prefer official pixel sizes (default 1024x1024, up to 4K); ratio forms 1:1 / 2:3 / 3:2 / 16:9 / 9:16 are legacy-only (not recommended). response_format is fixed to url. Returns a taskId; poll for the result. (The async channel endpoint is /api/cqt/generator/sora; an OpenAI-Images-protocol synchronous API is also available — see "Synchronous call".)
How to call
- Submit:
POST /api/cqt/generator/sora, returns a taskId. - Query:
GET /api/cqt/info/sora?id=taskId, poll until status = succeeded.
Capability map
POSTText to Image
Generate an image from text.
POSTImage to Image
Generate a new image from a reference + text.
POSTSynchronous call (GPT-Image)NEW
Generate an image synchronously via the OpenAI Images protocol — a single request returns the image (url or b64_json), no polling.
Common Errors
| Code | msg | Meaning | How to fix |
|---|---|---|---|
| 451 | content moderation | Prompt or image flagged | Edit the prompt or image and retry |
| 500 | timeout | Request timed out | Retry later; contact support if it persists |
Errors are returned as a bizCode, e.g. bizCode=451, msg=prompt_unsafe.
Tip: after saving your API Key via Authorize, every capability page supports "Try it".