The new feature: native GIFs and frame control
According to a report by IT之家 citing The Verge, ChatGPT can now create animated GIFs from a text prompt or a photo uploaded to the chat. The feature arrives together with OpenAI’s new Images 2.6 model and allows users to change the animation speed, work on individual frames and share the result with others.
The change is significant because it fills a previously stated limitation: before this update, ChatGPT could not generate GIFs natively in a single pass. In the example shared by the company, the system produces an animation from a request for a GIF of a dancing cat, also offering options to change the playback speed and view the content frame by frame.
Performance and rollout of the model
In addition to GIF creation, OpenAI states that Images 2.6 makes image generation faster. The company maintains that users currently generate more than 4 billion images per week through ChatGPT and its API, up from the 3 billion per week reported in September. These are figures stated by the company, not independently verified in the available material.
The Images 2.6 model is being rolled out to all ChatGPT and Codex users on web, desktop and mobile versions. OpenAI has also introduced the GPT-Image-2.6 Flare and Sunburst models into the API, presented as tools for faster, higher-quality image generation.
What remains to be verified
The material does not clarify the access limits for the GIF feature—for example, whether it is available on all plans or only to some users—nor does it specify output formats beyond GIF. No technical details are provided on how edited frames are handled, on file sharing, or on any usage restrictions.
No independent verification of the stated performance or of the weekly count of generated images is documented either. The available information is based on OpenAI’s communication and the source’s report, so claims about speed and volumes should be considered company statements pending external confirmation.
