THE GENERATION DESK

Shape the brief. Watch the image arrive.

Choose a model, write one focused creative brief and follow each server-confirmed state in real time.

Your result will open here.

The status rail follows the task from validation through processing, without polling the server.

Task progress

Created
Checking request
Generating image
Image ready
Generation failed
Request blocked

LIVE CODE PREVIEW

API examples

curl -X POST "https://chatgptimages.chat/api/generations" -H "Authorization: Bearer YOUR_CLERK_SESSION_TOKEN" \
  -H "Content-Type: application/json" \
  --data '{
  "model_name": "gpt-image-2.5-flare",
  "prompt": "Describe the image you want to create.",
  "reference_images": [],
  "idempotency_key": "A_NEW_UUID_FOR_EACH_LOGICAL_REQUEST",
  "challenge_token": "A_FRESH_TURNSTILE_TOKEN"
}'

These examples call this independent site’s API. They do not call OpenAI directly and never expose a provider key in the browser.

Shape the brief. Watch the image arrive.

Creative brief

Your result will open here.

Request paused