ChatGPT Images 2.6 Turns Prompts Into GIFs

OpenAI released ChatGPT Images 2.6 on Nov. 4, letting ChatGPT users turn a prompt or an existing image into an animated GIF, and shipped two matching API models, GPT-Image-2.6 Flare and GPT-Image-2.6 Sunburst. The company said the update brings "50% faster image generation," with Flare producing images "in as little as six seconds."

The launch was announced in a post on the OpenAI Developer Community forum and logged the same day in the API changelog. It arrives eight weeks after Images 2.5, which OpenAI released on Sept. 8.

Key takeaways

• ChatGPT can now create animated GIFs from a prompt or an image, with controls for playback speed and editing of individual frames.
• Users can save reference images of themselves, pets and characters and reuse them in new images without uploading them again.
• The API adds a "complexity" setting that defaults to high, plus SVG output and markup-guided edits; listed token prices match the 2.5 models.

ChatGPT GIF maker and saved reference images

"With ChatGPT Images 2.6, turn a prompt or image into an animated GIF. Adjust the speed, edit individual frames, and share it on repeat," OpenAI wrote in its announcement. Frame-level editing means a user can fix one bad frame instead of regenerating the whole loop.

The second consumer change targets a long-standing friction point. OpenAI said users can "keep reference images of yourself, pets, and characters ready to reuse in new creations—no need to upload them again." The announcement did not list which ChatGPT plans or regions get the features, or any limits on GIF length, frame count or resolution.

GPT-Image-2.6 Flare vs. Sunburst

The two API models keep the split OpenAI introduced with version 2.5. Flare is described on its model page as "image generation and editing for fast iteration." Sunburst is aimed at "detailed visual work," and OpenAI said it brings "higher-quality generation and greater consistency across edits."

According to the changelog, both models add adjustable complexity, SVG generation and animated GIF generation. Markup-guided editing, which lets developers mark up an image to steer an edit, works with both the 2.5 and 2.6 models.

The new complexity setting

OpenAI's image generation guide lists three complexity levels: low, medium and high. If a request leaves the setting out, both models use high. The guide cautions that "higher complexity can increase latency and introduce unwanted changes," and says the setting carries no separate fee. Rendering quality remains a separate "quality" parameter.

That default matters for budgets. Images are billed by output tokens, and one developer on OpenAI's forum reported that a medium-quality test produced 1,136 output tokens on 2.6 versus 344 on 2.5. OpenAI has not confirmed that comparison.

GPT-Image-2.6 pricing

The 2.6 model pages list, per 1 million tokens, $5 for text input ($1.25 cached), $8 for image input ($2 cached) and $30 for image output. Those rates match the 2.5 Flare and Sunburst rows on OpenAI's pricing page and are double the $4 input and $15 output image rates for GPT-Image-2. OpenAI's guide also notes that developers may need to complete API Organization Verification before using GPT Image models.

Context: a fast release cadence

Version 2.6 is the third GPT Image release in OpenAI's API changelog this year. GPT Image 2 arrived on April 21, followed by Images 2.5 on Sept. 8, which also promised lower latency and introduced the Flare and Sunburst names. Animated output pushes the product beyond still images into a format widely used in messaging and social media.

What's next

Developers should watch for OpenAI to clarify plan availability, GIF limits and per-image costs, none of which appeared in the announcement. Teams that move to 2.6 can set complexity explicitly rather than accept the high default, and compare output token counts against 2.5 before switching production traffic.

FAQ

Can ChatGPT make GIFs now?
Yes. With Images 2.6, ChatGPT can turn a text prompt or an uploaded image into an animated GIF, and users can change the speed and edit individual frames.

What is the difference between Flare and Sunburst?
Flare is built for fast iteration and can produce an image in as little as six seconds, according to OpenAI. Sunburst targets detailed work and more consistent results across repeated edits.

Did GPT-Image-2.6 get more expensive?
Listed per-token rates are the same as for 2.5. Actual cost depends on how many output tokens each image uses, which the new complexity setting can affect.

References

• Images, upgraded—introducing ChatGPT Images 2.6 — OpenAI Developer Community (Nov. 4, 2026)
• Changelog | OpenAI API — OpenAI
• Image generation | OpenAI API — OpenAI
• GPT-Image-2.6 Flare Model | OpenAI API — OpenAI
• Pricing | OpenAI API — OpenAI