OpenAI has introduced Images 2.6 across ChatGPT, Codex and its API. Its November 4 announcement adds GIF creation and image-editing controls; the GIF feature is rolling out in ChatGPT.
The big change
- Users can request looping GIFs or animate a sketch as the ChatGPT feature becomes available.
- API developers can choose Flare or Sunburst and adjust image quality separately from planning complexity.
- OpenAI claims faster generation and better consistency across edits; BIG CHANGE has not tested those claims.
Making an image move
OpenAI's release notes say users can adjust GIF playback speed and share it. Saved reference images can also be reused in later chats.
The announcement says Images 2.6 is available globally across ChatGPT and Codex subscription tiers. GIF creation is a rollout, not confirmation that every account already has it.
What changes for builders
The API guide describes GIF generation through the image-generation endpoint; the image-edit endpoint does not support GIF output. Developers can also use markup to indicate an area for editing and request SVG output. The guide says both 2.6 models support these features.
A separate complexity setting controls how much planning goes into a request, while image quality controls the rendering. For a builder, that separates the choice of a focused edit from the choice of visual detail. OpenAI positions Flare for fast iteration and Sunburst for more detailed work; those are vendor descriptions, not a benchmark conducted here.
