1. News
  2. Releases
  3. OpenAI · Images
ReleaseOpenAI · Images2 min read

ChatGPT can now turn a drawing or photo into an animated GIF

OpenAI's new Images 2.6 can animate a drawing or photo into a GIF from a plain text request, and it is live for every ChatGPT subscription tier.

By the Dashline deskSource: Marketing4eCommerce
ChatGPT can now turn a drawing or photo into an animated GIF
Photo: Santeri Viinamäki, CC BY-SA 4.0

OpenAI has released Images 2.6, an update to the image tool in ChatGPT. The headline feature is simple. Upload a drawing or photo, type what you want, and the tool turns it into an animated GIF. It is available worldwide, on every subscription tier, on web, mobile and desktop.

What changed

OpenAI says images are now created 50% faster. The update is live in ChatGPT and in Codex. The outlet Marketing4eCommerce notes that about 4 billion images are made every week, so speed matters at that scale.

The GIF tool works on brand mascots, characters, product photos, memes and illustrations. It can also turn a simple drawing into an animated character.

A real test

The outlet tested it with a hand-drawn giraffe. The only instruction was "please turn this drawing into an animation." ChatGPT replied that the giraffe gently moves its head, winks and smiles while showing off its pink bow, and that it kept the drawing's original style.

Why businesses may care

OpenAI says the model does a better job of keeping characters, products and brand elements the same when edits are made. That helps when you need several ads or posts that all look like they belong together. It also improves how elements are arranged in posters, marketing images and product pictures.

  • Local edits: change one part of an image without redoing the whole thing.
  • Better layout for advertising posters and product images.
  • SVG export: a vector format that stays sharp on websites, apps and print.

For developers

Developers can reach two new models through the API, which is how outside software connects to OpenAI's tools. GPT-Image-2.6 Flare favors speed and quick drafts. GPT-Image-2.6 Sunburst favors quality, detail and consistency, especially when editing.

Image markup, already in ChatGPT, is now coming to the API. A new setting also lets users choose how much effort the model puts into a request. Low suits quick edits. Medium and high are recommended for harder layouts. PicsArt, Higgsfield, Figma, fal and Sable were among the first to get access.

What is not known yet

The material does not give API pricing, usage limits for each subscription tier, or details such as GIF length and file size. Anyone planning to use it for client work or ads should test it on their own images first.

The takeaway

Try Images 2.6 on a logo, mascot or product photo you already have, and check that the animation keeps your brand look before using it in an ad.

More articles

You should really read these.

Next step

Let’s talk