OpenAI has added GIF creation to ChatGPT, so you can now build an animation from a text prompt or an uploaded photo, The Verge AI reports. The feature arrives with Images 2.6, OpenAI’s new image model. It’s rolling out to all ChatGPT and Codex users on web, desktop, and mobile.
It sounds like a small update, but it closes an odd gap. According to The Verge AI, ChatGPT couldn’t natively generate a GIF in a single step before this change. Now you ask for one and get the animation plus the controls to fix it.
1. GIFs start with a prompt or a photo
OpenAI’s own example is about as simple as it gets: “can you make a gif of a cat dancing?” ChatGPT returns the GIF, along with options to change the playback speed and view it frame by frame.
I think the photo option is the more useful of the two. A product shot or a team picture could become a short loop for a post or a slide, with no animation software involved.
2. You can fix it after it’s made
The editing tools are what stand out here. Once ChatGPT generates a GIF, you can:
- Adjust the animation speed
- Edit individual frames
- Share the finished GIF with others
Image generators have mostly worked like vending machines. You press the button, and if you don’t like what drops out, you press it again. Frame editing should mean you can fix one bad frame without rolling the dice on the whole animation. How well that works in practice is the thing to test.
3. Images 2.6 is faster, says OpenAI
OpenAI says the update speeds up image generation in ChatGPT. The report doesn’t include a figure for how much faster, so treat this as a company claim until people have run their own comparisons.
Speed matters more for animation than you’d think. If you’re going to iterate on a GIF, a long wait between attempts gets old fast.
4. Developers get two new API models
OpenAI is also launching GPT-Image-2.6 Flare and GPT-Image-2.6 Sunburst in its API, pitched for faster and higher-quality image generation. A two-model lineup usually signals a tradeoff between speed and quality, but the report doesn’t say which model does what.
That’s the part I’d want answered before building on either one. If you run image generation inside a product, cost per image and latency will matter more than the launch description.
5. The usage number may be the bigger news
OpenAI says users now generate over 4 billion images per week across ChatGPT and its API. In September, it reported 3 billion. That’s a jump of about a third, and it helps explain why OpenAI keeps investing in speed. Across billions of images, small speed gains add up.
GIFs also fit that growth pattern. They’re quick to make, easy to share, and every shared one is a small ad for ChatGPT. That last part is my read, not OpenAI’s, but it’s hard to ignore for a company whose image tools already spread through social feeds.
What we still don’t know
The launch leaves some practical questions open:
- Pricing: The report doesn’t cover API prices or whether usage limits vary by plan.
- Technical limits: There’s no word on maximum GIF length, resolution, or file size.
- Format limits: GIFs have no audio and limited color range, so this won’t replace video tools for polished work.
For social teams, newsletter writers, and anyone who needs a quick visual, this is a useful addition to a tool they may already have open. The next thing to watch is whether frame editing holds up on longer, more complex animations, or mostly shines on dancing cats. More details are available at The Verge AI.