ChatGPT can now make GIFs as OpenAI launches Images 2.6
OpenAI launched Images 2.6 on November 4, 2026, adding GIF generation to ChatGPT and Codex and two new image models, Flare and Sunburst, to the API. OpenAI says generation is 50% faster and users create 4 billion images per week.
- OpenAI's Images 2.6 adds GIF creation to ChatGPT and Codex, and two new models to the API.
- Marketers get quick animated mascots, social reactions and storyboards, plus SVG export and more consistent edits.
- Test GIFs and the API's complexity setting on small campaign assets before committing workflows.
Ask your own AI: ChatGPT Perplexity Claude AI Mode
On this page
OpenAI introduced Images 2.6 on November 4, 2026, and with it ChatGPT can now generate animated GIFs from a text prompt or an uploaded photo. The model is available in ChatGPT and Codex, and developers get two new API models. The announcement was published on OpenAI’s blog and covered by The Verge.
What does Images 2.6 change in ChatGPT and the API?
Images 2.6 adds native GIF creation to ChatGPT and Codex, and OpenAI says image generation is 50% faster than before. According to The Verge, ChatGPT previously could not produce a GIF in one step; now you can also change playback speed and edit individual frames. In the API, OpenAI offers GPT-Image-2.6 Flare (as little as six seconds at low quality) and Sunburst (higher quality and edit consistency at medium or high). OpenAI says Images 2.6 is available globally today across all subscription tiers, while Turkish outlet ShiftDelete.net describes a gradual rollout.
OpenAI also says users create 4 billion images per week across ChatGPT, Codex and the API, up from the 3 billion per week it reported in September, per The Verge.
Does ChatGPT GIF creation matter for marketers?
It lowers the cost of simple motion content, which is the practical point for marketing teams. OpenAI itself names animated brand mascots, social reactions and memes, and animated storyboards ahead of video production as use cases. Our read: this is useful for quick social assets and concept testing, not a replacement for a proper motion pipeline, and OpenAI has not published quality benchmarks beyond its own claims.
For developers, the API changes are more significant. Image markup (already in ChatGPT) comes to the API so edits can target specific areas, and SVG export makes outputs usable on websites, apps and print. A new complexity parameter controls how much planning the model does; OpenAI recommends low complexity for most edits and medium or high for scenes with many elements. Consistency of characters, products and brand details across edits is a stated focus, which matters for branded content at scale.
There is also a wider signal: image generation is becoming a mainstream use of AI assistants, and visual content created this way will increasingly appear in the channels you publish to. It continues the push we covered in OpenAI’s expanding web footprint with ChatGPT search, though this story is about creation tools, not search.
How to try Images 2.6 GIFs in your workflow
Start small and measure before you change any production process:
- Check whether Images 2.6 appears in your ChatGPT account; ShiftDelete.net notes the rollout may take a few days.
- Prompt a simple GIF, such as a mascot or product loop, then adjust speed and frames.
- If you use the API, compare Flare and Sunburst on your own assets and test low versus higher complexity for editing tasks.
- Review brand consistency across edits before publishing anything, and check OpenAI’s pricing page and system card for costs and safeguards.
- Watch file size and load impact if you place GIFs on web pages, since page speed still counts.