OpenAI's ChatGPT Hits 4 Billion Weekly Images as GIF Maker Rolls Out
Key Takeaways
- •Weekly image generation across ChatGPT and OpenAI's API has surpassed 4 billion, rising from the 3 billion reported in September.
- •Images 2.6 gives ChatGPT its first native GIF creation tool, producing animations from text prompts or uploaded photos with playback speed and frame editing options.
- •The update includes two new developer API models: GPT-Image-2.6 Flare, built for faster generation, and GPT-Image-2.6 Sunburst, aimed at higher-quality results.
- •The GIF feature is rolling out in stages, and Decrypt's testing suggests most users do not yet have full access.
- •The GIF tool arrives months after OpenAI shut down its standalone Sora video app in March, adding motion generation directly inside ChatGPT without a separate application.

OpenAI says users now generate more than 4 billion images per week across ChatGPT and its API, a figure up from the 3 billion weekly images the company reported in September. The milestone arrives as OpenAI rolls out Images 2.6, a new image model that adds animated GIF creation to ChatGPT. The Verge first reported the details of the update on November 4.
The jump of 1 billion weekly images — an increase of about a third — took place in roughly two months, the window covering the period between the launch of Images 2.5 and the rollout of the new 2.6 version. Images 2.5 arrived on September 8, the same month OpenAI reported the 3 billion weekly figure. Images 2.6 succeeds that release and extends across ChatGPT, Codex, and the company's developer API. Because the figure spans both the chatbot and the API, it reflects usage from everyday users and from applications built on OpenAI's models alike.
How the GIF Maker Works
ChatGPT users can now create animated GIFs from a text prompt or from a photo uploaded to the chatbot, giving the assistant a native animation tool for the first time. Prior to this update, ChatGPT could not natively generate a GIF in a single step.
Once a GIF has been created, users can change its playback speed, view the animation frame by frame, edit individual frames, and share the finished result with others.
OpenAI showcased the tool with the prompt “can you make a gif of a cat dancing?” The example included controls for animation speed and a frame-by-frame view of the resulting animation.
🚨 OpenAI just upgraded ChatGPT Images to 2.6! You can now create GIFs directly from prompts or images, use reference images, and control animation speed. Two new API models are coming too, promising 50% faster generation. Static AI images → Animated AI content. This is…
— Akshay Nandwana | DevRel & Android (@akshay81844) November 5, 2026
The rollout of the GIF feature appears to be gradual. Decrypt reported that the tool was not available on its accounts during testing. In Decrypt's test, ChatGPT produced a series of frames but not a finished animation, a result that suggests most users may not yet have full access to the feature. The phased approach means the tool may take time to appear for everyone.
Images 2.6 Brings Faster Generation
According to OpenAI, Images 2.6 enables faster image generation in ChatGPT, shortening the time it takes to produce an image. The model is rolling out to ChatGPT and Codex users on web, desktop, and mobile apps.
Developers are receiving two new API models: GPT-Image-2.6 Flare, which is built for faster generation, and GPT-Image-2.6 Sunburst, which is aimed at higher-quality results. Both succeed the original Flare and Sunburst models that debuted alongside Images 2.5.
OpenAI has not published a standalone announcement detailing the model's capabilities, and Decrypt said it could not independently test the new model's performance. For now, the public picture of Images 2.6 comes from third-party reporting and testing rather than OpenAI's own documentation.
The update follows ChatGPT Images 2.5, which brought sharper details, more precise editing, and generation speeds up to 50% faster than Images 2.0. That version also introduced the earlier iterations of the Flare and Sunburst API models, with Flare serving as the fast default and Sunburst built for premium editing precision.
In Decrypt's testing, Images 2.5 went head-to-head with Google's Nano Banana 2 across six categories, with each model winning three. The matchup ended even, though OpenAI's model still made mistakes when rendering text. No equivalent head-to-head against Images 2.6 has taken place yet, since Decrypt was unable to test the new version's performance.
Motion Generation Inside ChatGPT
The GIF feature arrives months after OpenAI shut down Sora, its standalone AI video app, in March. GIFs are simpler than the videos Sora produced, but the new tool adds motion generation inside ChatGPT without requiring a separate application. Unlike the standalone video app, the GIF tool produces short, looping animations directly within the chatbot.
OpenAI has not given a date for when every ChatGPT account will receive GIF creation. For now, Images 2.6 is reaching ChatGPT and Codex users while the GIF tool continues to roll out in stages. Whether the text-rendering errors that surfaced in Images 2.5 persist in the new model, and how the Flare and Sunburst API models hold up under independent testing, remain open questions as the rollout continues.
This article originally appeared on CoinCentral.