ChatGPT can now turn text prompts and photos into animated GIFs with OpenAI's Images 2.6 model
OpenAI has expanded ChatGPT's image capabilities with the release of its Images 2.6 model, which adds the ability to generate animated GIFs. Users can now produce short animations either by describing a scene in a text prompt or by uploading an existing photo as a starting point. Once generated, the output is editable - playback speed can be adjusted and individual frames can be modified, giving users a degree of control over the final result.
The key technical challenge in generating coherent animation is maintaining visual consistency between frames. If a character's appearance shifts noticeably from one frame to the next, the result looks unstable or broken. OpenAI says Images 2.6 handles this well enough to support flipbook-style animation, where the same subject remains recognizable across a sequence of images. This kind of frame-to-frame coherence has historically been difficult for diffusion-based image models, which generate each output somewhat independently.
The GIF format itself is a practical choice for this kind of feature. It is universally supported across browsers and messaging platforms, requires no video player, and loops automatically - making it easy to share without compatibility concerns. While the format has limitations compared to video (a restricted color palette, larger file sizes relative to quality), it remains widely used for short, looping visual content.
OpenAI also disclosed that users are now generating more than four billion images per week through ChatGPT, a figure that reflects how central image generation has become to the product since the rollout of native image tools earlier this year. The addition of animation marks a natural next step, moving the platform incrementally from static output toward richer visual media without requiring a full video generation pipeline.
