enzosoler.comreviews 08bench 9 devices
Published

ChatGPT now turns a prompt or a photo into a GIF with Images 2.6

Illustration: a strip of glowing film frames curling out of a laptop screen on a dark desk, each frame a slightly different dance pose of a simple white figure
Image: AI-generated illustration (Google Gemini)

OpenAI's Images 2.6 model, announced November 4, lets ChatGPT build an animated GIF from a text prompt or an uploaded photo, with playback speed and single-frame edits. It is rolling out on web, desktop and mobile, and reaches the API as GPT-Image-2.6 Flare and Sunburst. OpenAI has not published GIF limits or API prices.

If you make reaction GIFs, slide animations or quick product loops for Slack and decks, ChatGPT can now do it in one step. OpenAI announced its Images 2.6 model on November 4, and with it ChatGPT can turn a written prompt or an uploaded photo into an animated GIF. The feature is still rolling out, so it may not be in your account yet.

What changed

According to The Next Web, which credits The Verge, ChatGPT returns the GIF in the chat, and you can then change its playback speed, step through it frame by frame, edit single frames and share it. Before this update, ChatGPT could not produce a GIF natively in one step.

OpenAI’s Images 2.6 announcement also covers faster image generation in ChatGPT. Notebookcheck reports OpenAI’s claim as 50% faster with more consistent edits. OpenAI says people now make more than 4 billion images a week across ChatGPT and its API, up from 3 billion in September, per The Next Web.

Who gets it

Item What is reported
Surfaces ChatGPT and Codex on web, desktop and mobile (The Next Web)
Plans All ChatGPT and Codex users, rolling out (The Next Web); Notebookcheck says OpenAI includes the free tier
GIF limits Not published by OpenAI (Notebookcheck)
API models GPT-Image-2.6 Flare for speed, GPT-Image-2.6 Sunburst for quality
API prices Not listed in the coverage

Notebookcheck also reports two details the other outlets don’t mention. You can save an uploaded photo under a name and call it in later chats with an @ mention, such as a pet photo saved as “@Pepper.” And ChatGPT cannot edit a GIF you upload; it only animates stills or prompts. OpenAI’s help page, per Notebookcheck, does not say where saved photos are managed or how to delete them.

What to do

  1. Try it on a throwaway idea first. Ask for a short loop, then adjust the speed or a single frame before you ask for a full redo.
  2. Keep real people out of saved references. A named, reusable photo plus animation makes it easy to put a coworker or customer into a scene they never agreed to. Until OpenAI documents how to delete saved photos, save only objects, pets or your own face.
  3. Check your company’s AI policy before using it for client work. Generated images are stored in your ChatGPT account, and Notebookcheck says they are removed only when the conversation that made them is deleted.
  4. Developers: wait for prices. Flare and Sunburst are named, but no USD rates were in the coverage at publication. Don’t swap a production image pipeline until OpenAI’s pricing page lists them.

Work GIFs move from a separate tool into ChatGPT

Most teams that make GIFs for docs, release notes or social posts use a screen recorder or a design tool. Folding the job into ChatGPT, at no extra charge on plans people already pay for, removes a step for quick visuals. That is our inference from the rollout: OpenAI has not said anything about replacing other tools, and with no published limits, heavy users can’t yet plan around it. For the other recent change to ChatGPT’s paid plans, see ChatGPT Pro for Work and doubled Codex limits. For OpenAI’s latest model pricing, see the GPT-6.1 Terra price cut.

One email when a review goes up. Unsubscribe from any of them.