ChatGPT Images 2.6 is OpenAI's latest image model in ChatGPT, Codex and the API. Turn a prompt or image into an animated GIF, adjust its speed and edit frames. Save reference images of yourself, pets or characters and reuse them anytime. Generation is 50% faster, with stronger consistency across edits. Developers get GPT-Image-2.6 Flare (as fast as six seconds) and Sunburst (higher quality), plus image markup and SVG export.

Hey PH community 👋

Excited to hunt ChatGPT Images 2.6 today.

OpenAI just upgraded image generation across ChatGPT and Codex, and the two new API models generate images 50% faster than before while keeping characters, products and brand details more consistent as you edit. OpenAI posted a short look at it on X here.

Here's an example from the community:

https://x.com/Dheepanratnam/status/2118138765487100027

Dheepan ran some quick GIF tests with Images 2.6, going from 9 frames up to 24fps.

GIFs are the part I'm most excited about. You can turn a prompt or an image into an animated loop, adjust the speed and edit individual frames, which makes brand mascots, reactions and storyboards something you can make right inside ChatGPT.

Reference images are the other one I'd test first. You save yourself, your pet or a character once and reuse them in new creations without uploading them again every time.

For builders, GPT Image 2.6 Flare and Sunburst are coming to the API, with Flare generating images in as little as six seconds and Sunburst focused on higher quality and consistency, plus image markup and SVG export.

You can try it in ChatGPT right now: https://chatgpt.com/images

Give it a spin and let the team know what you create with ChatGPT Images 2.6 🚀

PS: Tagging @sama and @greg_brockman if you're here for any additional comments/ thoughts.

GIF generation is a pretty cool addition. Animation is definitely fun to play around with.

It is already doing well visually. And I thought that there is no room for improvement left. And they found it :)