OpenAI: Images 2.6 adds GIF generation and SVG export to ChatGPT
OpenAI released Images 2.6 on November 4, 2026, adding animated GIF output, SVG export, a reusable reference library, and image markup to ChatGPT’s image generation. The same features are available in the API through two new models, GPT-Image-2.6 Flare and GPT-Image-2.6 Sunburst.
The GIF feature generates a short looping animation from a text prompt or an uploaded photo. Within ChatGPT, you can adjust playback speed and edit individual frames before exporting. The output is built from related still images rather than a continuous motion model, which works well for short loops, reactions, and stylized motion but produces artifacts on complex movement or multi-subject scenes.
SVG export produces vector output for logos, icons, and simple illustrations — removing the step of tracing generated images by hand.
The reference library stores characters, mascots, pets, or people so they can be reused across prompts and sessions, which helps keep a character’s appearance consistent across storyboards and concept work.
Image markup lets you circle or annotate part of a generated image to show which region should change, rather than describing the edit only in text.
Why it matters for designers
The combination of GIF generation and per-frame editing brings short motion work into ChatGPT without a separate tool. SVG export and the reference library address two frequent friction points in AI-assisted visual work: getting editable output and keeping characters consistent. GIF rollout is still gradual, so not all accounts have it yet.