ChatGPT Images 2.6 (2026): The Best New Way to Make GIFs With AI
OpenAI has given its image generator its biggest upgrade of the year. ChatGPT Images 2.6, announced on November 4, 2026, turns a simple prompt or an existing photo into an animated GIF, remembers the people, pets and characters you want to reuse, and ships with two new models for its API. In short, it is the moment AI image generation stops being only about still pictures.
What Is New in ChatGPT Images 2.6
The headline feature is GIF creation. Describe a scene in words, or upload a photo, and ChatGPT Images 2.6 can animate it into a short looping clip. You can adjust the playback speed, and you can edit individual frames when a detail does not look right, rather than starting the whole image again. For social posts, chats, presentations and quick explainers, that turns a static picture into something that moves, without opening a video editor.
The update also adds saved reference images. If you have a character, a person or a pet you want to appear again and again, you can save their photo as a reference and reuse it across generations. The idea is consistency: the same face, the same dog or the same invented character, kept recognisable from one image to the next, instead of drifting every time you start a new prompt.
There is a smaller but useful addition for designers too: SVG export. Images from ChatGPT Images 2.6 can now be saved as scalable vector graphics, which stay sharp at any size and are far easier to drop into logos, icons, slides and illustrations than a fixed-size raster file.
Two New API Models: Flare and Sunburst
Developers get two new API models behind ChatGPT Images 2.6. GPT-Image-2.6 Flare is built for speed, producing images in as little as six seconds at low quality, while GPT-Image-2.6 Sunburst is aimed at detail and edit consistency, holding a subject steady across repeated edits. OpenAI says image generation is 50 percent faster overall, although that is the company’s own claim rather than a guarantee for every job: its published Flare benchmark shows a drop from 7.08 seconds to 4.84 seconds, which is about 32 percent. Either way, the direction is clear, and faster generations make the new GIF tools practical to use.
Availability: Who Can Use ChatGPT Images 2.6
ChatGPT Images 2.6 is available in ChatGPT and in Codex, on all tiers, across web, desktop and mobile. There is one catch worth knowing before you go looking for the GIF button: GIF creation is rolling out in stages, and some accounts do not have it yet. If it is missing from your account, it is not a fault with your app; OpenAI is still switching it on account by account. The image model itself, the saved references and the SVG export should reach users more quickly than the GIF tools.
The scale behind the ChatGPT Images 2.6 launch is striking. ChatGPT and the API now generate more than four billion images a week, up from three billion in September, according to OpenAI. That growth helps explain why the company is investing so heavily in speed: every second saved per image is multiplied across billions of generations.
Why This Update Matters
For creators, the appeal is obvious. A GIF made from a photo in seconds is a faster way to make a post stand out, mock up a storyboard, or explain a product, and reusable references mean a brand mascot or a story character can stay consistent across a whole campaign. For everyone else, the release is another sign of how quickly AI tools are becoming part of everyday work. As we reported, AI assistants are already changing how people work, and image generation is following the same path from novelty to daily habit. ChatGPT Images 2.6 also raises the bar for what people expect from an image tool: animation and consistency are now part of the package, not extras.
How to Try the New GIF Tools
For most people, getting started is simple. Once ChatGPT Images 2.6 is live on your account, open ChatGPT on the web, desktop or mobile and ask for an image as you normally would, then look for the animation option in the image tools. A good first test is a photo with a clear subject and a little movement in it: a pet looking at the camera, a person waving, or a landscape with clouds or water. Describe the motion you want in plain words, such as asking for a slow pan across the scene or for the dog to tilt its head, and let the model produce a first draft before you fine-tune it.
Once the GIF exists, the frame editor is where the real control sits. If a hand bends the wrong way or a face shifts between frames, you can correct that single frame instead of regenerating the whole clip, which saves both time and credits on paid plans. Creators making social content should start with short loops of two to four seconds, since those play smoothly in feeds and chats and are small enough to share anywhere. Longer, more detailed animations will look better but take longer to generate, especially while Flare is the speed-first model and Sunburst is tuned for detail and consistency.
Saved references deserve a little planning, too. Pick one clear photo for each character, pet or person, save it once, and reuse that same reference for every new scene. Jumping between different reference photos of the same subject is the fastest way to lose the consistency the feature is meant to provide. Designers, meanwhile, should try the SVG export early: a logo or icon generated and exported as a vector can be resized for a website, a slide deck and a printed poster from the same file, without the soft edges that come from enlarging a raster image.
If your account already has the update, the best way to judge it is to try the GIF tool on a photo you know well and see how faithfully it moves. For the full rundown of the announcement, see TechRepublic’s coverage of the ChatGPT Images 2.6 launch. If the GIF option has not appeared for you yet, check back over the coming days as the rollout continues.
