About these notes

Notes kept while building the input side of an independent browser interface for OpenAI's GPT Image 2.5 — sketch mode, the prompt rewriter, and the comparison work that came with a generation change.

The product is GPT Image 2.5 Image Generator. It runs both models OpenAI ships and prints the model ID the API returned under every result, which is the piece consumer chat products leave out.

Disclosure

I build the product referenced here, so weigh the notes accordingly. It is an independent interface, not affiliated with or endorsed by OpenAI. Figures quoted were measured against a live endpoint; where something is an estimate, the sentence says so. Nothing here is sponsored and there are no affiliate links.