OpenAI Releases Images 2.6 With Animated GIFs

OpenAI released Images 2.6, a new version of its image-generation model. OpenAI is the company behind ChatGPT. The model is live now in ChatGPT, in Codex (OpenAI's coding tool), and through its API, on all pricing tiers. The headline new skill is animation: a photo, a simple sketch, or a text prompt can become a short looping GIF (a tiny animated image file) — a brand mascot come to life, a social-media reaction, or a rough storyboard for video. Plain image generation is also 50% faster than before. The API gets two variants: Flare, a fast option that returns a low-quality image in six seconds, and Sunburst, a more precise option for edits that keeps a character's or product's details intact. OpenAI says ChatGPT, Codex, and the API together now produce four billion images a week — over five hundred million a day.

None of this touches coding agents directly. The model handles pictures and video frames, not reasoning or working inside a code repository, so it changes nothing about the agent stack people build software with. For content work, the time savings are real and concrete. Animating a character or cutting a short social clip used to mean building it frame by frame in a video editor; now it's one prompt and a finished looping GIF. Where image and video generation already run through AI agents, a 50% faster model just means less waiting between iterations — not a jump in the quality of the image itself.