ChatGPT Adds Animated GIF Creation Alongside the Images 2.6 Rollout
OpenAI is rolling out a feature that lets ChatGPT create animated GIFs from a text prompt or an uploaded photo, shipping alongside its new Images 2.6 model. The feature is rolling out, not yet available to every user.
The GIF tools go beyond generation. According to The Verge, users can adjust the playback speed of an animation, edit individual frames, view a GIF frame by frame, and share finished GIFs with others. Those controls turn a one-shot generation into a short editing workflow. The output becomes a sequence of addressable frames (each one can be reached and changed on its own) rather than a single image.
Images 2.6 is the model underneath the rollout. OpenAI says it enables faster image generation in ChatGPT, and it is reaching all ChatGPT and Codex users across web, desktop, and mobile apps. In the API, OpenAI is launching two variants, GPT-Image-2.6 Flare and GPT-Image-2.6 Sunburst, which the company positions as faster and higher quality for developers building image generation into their own products.
The existing workload is large. Per the same report, users now generate more than 4 billion images per week across ChatGPT and OpenAI's API.
The feature has a visible lineage. ChatGPT Images 2.0 launched on April 21, 2026, with improved text rendering, multilingual support, and visual reasoning, according to OpenAI's launch post. OpenAI's release notes list Images 2.0 as available on all ChatGPT plans. OpenAI's DevDay 2026 recap, published on September 29, described more than 20 major announcements across ChatGPT, Codex, models, and new forms of working. The GIF feature and Images 2.6 sit within that sequence of releases.
The broader context here is the move from still images to time-based output. A GIF is a bounded sequence of frames, each shown for a set duration on a timing schedule. Exposing per-frame editing means the generation pipeline has to hold visual consistency across frames, not just within one canvas. The questions that follow are the usual ones for generative media: how frame coherence holds up over longer loops, how much latency and compute a multi-frame output adds relative to a single image, and whether the faster Images 2.6 path lowers those costs enough to make animation practical at the volume OpenAI already handles for stills.
In my view, the more interesting consequence is distribution rather than generation. Animated GIFs are a shared format, used in messaging threads, social feeds, and support channels, and a tool that makes them from a prompt or a family photo lowers the effort to almost nothing. That helps people who previously needed a design tool and an afternoon to make a looping clip. It also means synthetic animation will circulate in the same places as ordinary GIFs, where provenance is already thin. The sharing control is where those two things meet, and attribution and labeling are worth watching as the rollout reaches more users.
Taken together, the announcement extends ChatGPT's image pipeline into frame-level editing and sharing, with Images 2.6 as the foundation and a developer-facing API variant alongside it. For a product already producing billions of images a week, the practical test is whether animation becomes a routine output for ordinary users rather than a novelty.
