Overview
- OpenAI began rolling out Images 2.6 on Wednesday, bringing a built-in GIF workflow that turns text prompts or uploaded photos into looping animations.
- The GIF tool creates short loops of 2–16 frames (default nine) at roughly 200 milliseconds per frame and lets users change playback speed, edit individual frames, and share or download the result.
- Developers get two GPT-Image-2.6 API variants—Flare for faster generation and Sunburst for higher quality—but GIF output is implemented in the code behind a feature flag and lacks full public documentation today.
- OpenAI says weekly image generation has climbed to about 4 billion across ChatGPT and its API, and Images 2.6 also promises faster image creation and more consistent editing than prior releases.
- The feature is aimed at quick reaction loops and simple motion rather than full video production, and its gradual rollout means some users see finished GIFs while others only get generated frames or must wait for broader access.