Images 2.6 logoImages 2.6
Back to home

What Is Images 2.6?

A plain guide to OpenAI’s Images 2.6 model: what’s new, where you can use it, and how to try it.

Last updated: 2026-11-06

Images 2.6 is the image model update OpenAI released on November 4, 2026. It adds animated GIFs, generates images faster, and keeps details steadier across rounds of edits. You can use it in ChatGPT, in Codex, through the API, and in the generator on this site, which is independent and not affiliated with OpenAI.

Images 2.6, GPT Image 2.6 or ChatGPT Images 2.6?

All three names point to the same release. OpenAI calls the update Images 2.6, and people often say ChatGPT Images 2.6 when they mean using it inside ChatGPT. GPT Image 2.6 is the name developers see, because the API models are called GPT-Image-2.6 Flare and GPT-Image-2.6 Sunburst.

Release date and where you can use it

OpenAI released Images 2.6 on November 4, 2026. It is available in ChatGPT and Codex on every subscription tier, on web, desktop and mobile. Developers reach it through the OpenAI API. OpenAI also says its products now make 4 billion images a week.

What’s new in Images 2.6

  • Animated GIFs. The model can make short loops of characters, mascots, product shots, sketches, reactions and storyboards. In ChatGPT, a prompt or an uploaded photo can become a GIF, and you can change the playback speed and view and edit single frames.
  • Faster generation. OpenAI says images generate 50% faster than before, and that Flare can produce an image in as little as six seconds at low quality. Its guide also says complex prompts may take up to 2 minutes. These are OpenAI’s figures, and we haven’t measured them.
  • Markup edits in the API. Image markup lets you point at the exact area to change. It was previously only in ChatGPT.
  • Consistent edits. Characters, products and brand details hold steadier across rounds of edits.
  • SVG export. The API can return vector files for websites, apps and print.
  • A complexity control. A new parameter sets how much planning the model does before drawing, with no separate fee. Low suits quick edits. Medium or high suits scenes with many elements or constraints. On this site the setting is called Planning.

ChatGPT also lets you save references of yourself, pets and characters and reuse them. That library is a ChatGPT feature and isn’t part of this site.

How it compares with the previous version

If you’re weighing it against the release before it, the short list is this: GIFs are new, OpenAI reports images arriving 50% faster, and image markup, SVG export and the complexity control have reached the API. Readable text in posters and layouts is a stated strength of Flare, though text won’t be right on every run.

Flare and Sunburst in brief

Developers choose between two models, and so do you on this site. Flare is for speed and professional layouts like posters and product imagery. Sunburst is for more detail and more consistent edits at medium or high quality. They share the same token rates at OpenAI. The full comparison is in our Flare vs Sunburst guide.

Is it free on this site?

In ChatGPT and Codex, OpenAI offers it on every subscription tier. On this site you get 20 welcome credits when you sign up, which covers 4 images at 1K, and daily check-ins add more. After that you can buy a credit pack or a monthly plan, starting at $9.90.

How to try it on this site

  1. Open the generator and sign in with Google.
  2. Choose a mode: Image, Edit, GIF or SVG.
  3. Pick Flare or Sunburst, then set the quality and planning level. The other options depend on the mode, such as aspect ratio and resolution for images, or frames and speed for GIFs.
  4. Write your prompt. In Edit mode, upload 1 to 4 images first, and mark the area to change if you want to.
  5. Press Generate, then download your PNG, GIF or SVG. Your last 12 runs stay listed under the result.

Which mode to start with

If you’re new to the model, start with the Image tab and a plain 1K run at 5 credits, because it shows you how Flare and Sunburst read your prompt. Move to the GIF tab when you want motion, and to the Edit tab when you already have a photo and only one part of it needs to change. The SVG tab suits logos and icons you’ll scale up later.

Common questions

Can Images 2.6 make GIFs?

Yes. GIFs are the headline addition. On this site a text prompt becomes a looping GIF of up to 16 frames, and our GIF maker guide covers prompts, frames and speed. Photo-to-GIF and frame-by-frame editing are available in ChatGPT but not here.

Does Images 2.6 make video?

This site gives you PNG, GIF and SVG files. For motion, you get a short looping GIF, not a video file.

Does it support transparent backgrounds?

Yes. In Image mode here, tick Transparent background and you get a PNG. At 1K it costs 5, 8 or 14 credits at Low, Medium or High quality.

What resolution does it output here?

Images and edits come out at 1K, 2K or 4K. At 16:9 that means 1536x864, 2048x1152 or 3840x2160. GIF frames are 512x512.

Which quality levels can I use here?

Low, Medium and High. OpenAI’s API also lists xhigh and max, which this site doesn’t offer.

Is this site run by OpenAI?

No. This is an independent site, not affiliated with, endorsed by or sponsored by OpenAI. “ChatGPT” and “OpenAI” are OpenAI’s trademarks.

Sources