How to make a product GIF with ChatGPT Images 2.6

OpenAI announced Images 2.6 on November 4, 2026. The product post says it is in ChatGPT, Codex, and the API, and that you can bring motion to images with GIFs, including turning a product shot into a loop. This page sticks to what OpenAI has published. It is not a feature list we tested ourselves.

What OpenAI has published

The announcement says Images 2.6 generates images 50% faster than before, and that the API offers two models: GPT-Image-2.6 Flare and GPT-Image-2.6 Sunburst. OpenAI says Flare can generate images in as little as six seconds when quality is set to low, and that Sunburst is for greater detail and more consistency across edits at medium or high quality. The same post says Images 2.6 is available globally in ChatGPT and Codex on all subscription tiers. It points to a separate page for API pricing rather than listing a per-GIF price.

OpenAI’s community announcement for the launch also describes GIFs inside ChatGPT: turn a prompt or an image into an animated GIF, adjust the speed, and edit individual frames. It says you can save reference images of yourself, pets, and characters and reuse them. Treat that as a description of the ChatGPT product. The API guide is more specific, and it does not say the same thing in the same words.

What the API guide says about GIFs

The image generation guide says both Flare and Sunburst support animated GIF output through the Image API. You set output_format to gif on POST /v1/images/generations and describe the subject, the motion, and the loop in the prompt. Documented controls are frame_count from 2 to 16 (default 9) and frame_duration_ms from 20 to 2000 milliseconds per frame (default 200, rounded to 10 ms). Size can be auto, or width and height that are positive multiples of 16. The guide says GIF output is for generation. The image edit endpoint does not support GIF output.

That last line matters for a listing. If your goal is “take this exact packshot and move it,” the published API path for GIFs is a new generation from a prompt, not an edit of the file you already use on the product page. ChatGPT’s own announcement talks about starting from an image. Those are different surfaces. Check the current screen in ChatGPT before you assume an uploaded photo becomes the first frame.

A careful way to try it on a product

• Start from the photo you would actually publish, or from a prompt that names the product, the background, and the motion.
• Ask for a short loop and no new words on the packaging. Generated letters on a label are a common place these tools drift.
• If you are in the API, follow the guide: generations endpoint, output_format gif, and a frame count you can preview. Do not expect the edits endpoint to return a GIF.
• Save the file and open it next to the original photo. If the bottle shape, color, or printed label changed, do not put it on the listing.
• Check the marketplace rules before you upload. A GIF is not the same slot as a Shopify or Etsy video.

When a dedicated tool is the simpler path

ChatGPT is a general image product. Motionframe is only the listing step: upload the photo, pick a catalog motion, and download an MP4 and a GIF. The clip is meant to start from that photo, in 1:1, 4:5, or 9:16, for 3 or 5 seconds. If the photo is already the source of truth, that is the shorter path. If you want a new scene, a mascot, or a storyboard, Images 2.6 is what OpenAI is describing.

• OpenAI: Introducing Images 2.6
• OpenAI API: image generation guide
• GPT-Image-2.6 Flare model page
• OpenAI community announcement
• OpenAI post on X
• Make a listing clip from a photo
• Images 2.6 and Motionframe, side by side