Hm, which tier are you on? See the rate limits at the bottom of this page: https://platform.openai.com/docs/models/gpt-image-1. (E.g., Tier 1 supports 5 images a minute.)
3 Likes
OpenAI has extended the requirement for ID-based verification to all use of this AI model modality. You will need to in to “general”, consent to giving a third-party company uploads of your government ID, other demanded documents, make videos of yourself from multiple angles - biometrics that cannot be reset after a “hack”. You will find it in organization->general. However, the whole Playground page will be denied if you have not done this.
Then other topics have other orgs reporting they verified, yet still get an API error from the model (you’ll need to make RESTful calls until there is an updated SDK - OpenAI put refusal of unknown models right into the code).
Extra tip: see if the extent of your “images” models in your org limits is entirely older models. Attempts to play around with generating new API keys won’t bypass this:
