Overview
- OpenAI began a gradual rollout of its Images 2.6 model this week, adding native GIF generation inside ChatGPT and the Codex coding tool on web, desktop, and mobile for all subscription tiers.
- The GIF tool turns text prompts or uploaded photos into short looping animations and provides playback speed controls plus frame-by-frame inspection and editing.
- A code change in OpenAI’s Codex repository shows the Images API can output 2–16 frame animations with a default of nine frames at about 200 milliseconds per frame, but that API GIF support is gated behind feature flags and lacks public documentation.
- Developers received two Images 2.6 API variants—GPT-Image-2.6 Flare for speed and GPT-Image-2.6 Sunburst for quality—while the GIF feature is explicitly limited to short loops rather than replacing full video tools.
- OpenAI says image generation has climbed above 4 billion images per week, a growth signal for creators and social media, and broader access and API documentation remain the next things to watch.