Make a photo square with AI
Re-renders a photo as a square without cutting anyone out of it: the people and the setting are kept and the rest of the frame is composed to fill the square. A photo that is already square is rendered again, so calling twice gives a second rendering. Requires a bearer session.
The square comes back as base64 and nothing is stored. Upload it through the normal image upload to use it as a cover.
Shares the limits of POST /images/generate: 10 an hour and 30 a day per user; past either limit the answer is 429 with Retry-After. Errors from the image provider are not passed through: a photo the provider refuses answers 400, a temporarily unavailable provider 503, and any other provider failure 502, each with a generic message.
Authorizations
In the browser, authentication rides on the httpOnly session cookies set by /cognito/signin or the Google / Apple sign-in at /cognito/oauth/{provider}. For scripts and for Swagger UI testing, paste an API key (fh_live_…, created with POST /api-keys); a Cognito JWT is accepted too. Sign-in does not return a token in its body.
Request Body
Responses
The square photo