OpenAI has released Images 2.6, an update to the image generator inside ChatGPT that can turn a text prompt or a still photo into an animated GIF.

Announced on 4 November, the update is available in ChatGPT and the Codex coding tool, and through the API, on every subscription tier including Free, across web, desktop and mobile. OpenAI says image generation is about 50% faster than before, and that users can adjust a GIF's playback speed and edit individual frames.

For developers, the update arrives as two new models: GPT-Image-2.6 Flare, tuned for speed and able to return a low-quality image in as little as six seconds, and Sunburst, tuned for detail and consistency. The API also gains SVG output and new editing controls. Flare suits quick drafts and poster or product layouts, while Sunburst is the steadier choice for detailed, consistent images.

There is a catch. GIF creation is still rolling out gradually. OpenAI's help pages said on 7 November that the feature was still rolling out, so some accounts may not see it yet. Users can also save photos as references for later chats.

OpenAI says its products now produce 4 billion images a week across ChatGPT, Codex and the API, a scale that helps explain why faster generation matters to the company.