ChatGPT Adds Animated GIF Creation as OpenAI Rolls Out Images 2.6

Until now ChatGPT could not produce a GIF natively in one step. The feature depends on the model keeping a subject consistent from one frame to the next, which earlier image generators struggled to do. In an example the company shared, the prompt "can you make a gif of a cat dancing?" produced a series of slightly varied images of the same character, played in sequence like a flipbook, with controls to slow the animation down or step through it frame by frame.

The Decoder said the model's frame-to-frame consistency appears strong enough to keep the character recognisable, which earlier OpenAI models would probably not have managed.

OpenAI says Images 2.6 also makes image generation faster in ChatGPT. The update is rolling out to all ChatGPT and Codex users on the web, desktop and mobile apps.

For developers, the company is releasing two API models: GPT-Image-2.6 Flare for faster generation and GPT-Image-2.6 Sunburst for higher-quality output. The split follows the pattern OpenAI set with Images 2.5 in September, when it positioned Flare as the default for most applications and Sunburst as the premium option, with longer generation times and tighter control over edits. The reports did not include pricing for the new API models.

The release also shows how fast image generation has grown as a use of OpenAI's products. The company says people now create more than 4 billion images a week across ChatGPT and its API, up from the 3 billion a week it reported in September with the Images 2.5 launch. Those figures come from OpenAI and cover all image activity, not just GIFs.

For marketing and social media teams, the practical change is that short looping animations, previously a job for design tools or video generators, can now be made and adjusted in the same chat window used for drafting copy and still images. The reports did not say whether GIF output is yet available to developers through the API.