On November 4, 2026, OpenAI rolled out Images 2.6, the new image model inside ChatGPT and Codex — and it turns the chatbot into a GIF maker. Type a prompt or upload a photo, and instead of a still image you get a looping animation. You can then change how fast it plays, edit individual frames, and share the result, all without leaving the chat.
What is new: GIFs, frame by frame
The headline feature is one-step animated GIF creation. Until now, ChatGPT could generate still images; making them move meant exporting to another tool. Images 2.6 closes that gap: a dancing cat from a short text prompt was OpenAI's own demo example. After generation, users can adjust playback speed, revise specific frames, and inspect frames one by one before sharing. The update also adds reusable reference images, so a character, mascot, or product shot can stay consistent across generations — handy for reaction images, simple storyboards, and social posts.
Faster images, two new API models
For developers, OpenAI is shipping two new API models: GPT-Image-2.6 Flare, tuned for speed, and GPT-Image-2.6 Sunburst, tuned for higher-quality output, edit consistency, and SVG export. OpenAI says image generation is 50% faster overall than the previous generation, and that users now create more than 4 billion images a week across ChatGPT and the API, up from over 3 billion in September. Those volume figures are OpenAI's own counts, not independent measurements.
Availability and pricing
Images 2.6 is rolling out globally across ChatGPT and Codex on the web, desktop, and mobile, and it covers every subscription tier — including free users, not just paying ones. There is no separate charge inside ChatGPT; it arrives as part of the existing plans. The rollout began gradually on November 4, so if the GIF option has not appeared in your app yet, it should arrive as the rollout reaches you.
The verdict
This is a smaller, more practical upgrade than a new flagship chatbot — and that is the point. GIFs are the internet's native language for reactions, memes, and quick explainers, and ChatGPT just learned to speak it. The frame-editing controls matter more than the novelty: being able to fix one bad frame instead of regenerating the whole clip is what makes this usable rather than a gimmick. The thing to watch is quality over time — short loops are forgiving, but longer, detailed animations will test the model. For now, if you make social content, try animating a photo you already have; that is where the feature shines first.
Comments
Post a Comment