Social Creators
Develop animated reactions, memes and expressive characters for posts and conversations.
GPT Image 2.6 generates images, animated GIFs and SVG designs from your ideas. Create expressive motion, refine visuals with image markup, and keep characters and details more consistent across edits.
Explore animated characters, scalable SVG designs, imaginative scenes, and guided edits. Use a starter prompt to create your own variation.

GPT Image 2.6
OpenArt: A Falling Market

GPT Image 2.6
OpenArt: A Rising Market
GPT Image 2.6
Field & Flour: SVG Bakery Logo

GPT Image 2.6
Sketch to Character

GPT Image 2.6
Picsart: Glass Whale

GPT Image 2.6
Markup-Guided Room Edit

GPT Image 2.6
Sable: A Miniature Product World
OpenAI’s image model family expands into animation and vector output.
OpenAI released Images 2.6 on November 4, 2026. Also known as ChatGPT Images 2.6, the family creates and edits visuals from text and image inputs, with new GIF creation and SVG output capabilities.
Two versions serve different needs: Flare emphasizes faster iteration and improved layouts, while Sunburst emphasizes detailed compositions and consistency through repeated edits.
The release supports three connected creative tasks:

Create motion and vectors, then refine your images with more control over detail and composition.
Turn an idea into a moving reaction, character or visual loop. GPT Image 2.6 introduces GIF creation, with launch examples ranging from animated sketches to expressive mascots.
Bring generated designs into website, app and print workflows with SVG output. This vector format adds a new way to carry an idea beyond a raster image.
Refine a character, product or composition while preserving more of its defining details. Sunburst improves consistency across editing rounds, particularly at medium and high quality.
Use sketches and image annotations to indicate where a change belongs. Spatial guidance makes your intent clearer, without promising that every unmarked pixel stays unchanged.
Flare focuses on fast iteration and improves layouts for posters and product imagery. OpenAI reports generation in as little as six seconds at low quality; actual time depends on the request and settings.
Quality controls visual fidelity, while complexity controls planning effort. Lower complexity suits focused edits; more involved scenes can benefit from additional planning.
The model serves creators working with still images, animated characters and vector designs.
Develop animated reactions, memes and expressive characters for posts and conversations.
Explore image compositions, refine details and produce SVG designs for digital layouts and print.
Create campaign imagery, animate mascots and preserve recognizable product or brand details as you revise visuals.
Turn sketches into animated characters and explore movement through short visual sequences before a larger production.
Turn your idea into a visual, then refine the details with GPT Image 2.6.
Enter a prompt describing the subject, composition and style. Add a reference image when you want to guide an edit.
Select the model version and adjust the available output settings to suit your project.
Review the displayed credit cost before generating. Inspect the result, download it or use it as a reference for another edit.
Explore GPT Image 2.6 inputs, output capabilities and creative controls.
| Developer | OpenAI |
| Release date | November 4, 2026 |
| Model versions | Flare and Sunburst |
| Inputs | Text and images |
| Image workflows | Image generation and instructed editing |
| New output capabilities | Animated GIFs and SVG |
| Editing guidance | Sketches and image annotations |
| Creative controls | Separate quality and complexity controls |
| General video and audio | Not supported; GIF animation is a separate capability |
| Maximum raster resolution | Not confirmed in the reviewed 2.6 documentation |
Last updated: November 5, 2026
Compare the new release with its predecessor using the capabilities described in the release documentation.
| Spec | GPT Image 2.6 | GPT Image 2.5 |
|---|---|---|
| Versions | Flare and Sunburst | Flare and Sunburst |
| Text and image inputs | Supported | Supported |
| Image generation and editing | Supported | Supported |
| GIF creation | Added with this release | Not listed |
| SVG output | Added with this release | Not listed |
| Editing consistency | Improved against corresponding 2.5 baselines | Baseline for the 2.6 comparison |
| Separate planning control | Complexity control introduced | Not listed |
GPT Image 2.6 Advantages: Adds GIF and SVG workflows, with faster iteration and stronger detail preservation through edits.
GPT Image 2.5 Advantages: Supports text-driven image creation and reference-based editing, with a familiar still-image workflow.
Explore GPT Image 2.6 for animation, vector output and iterative visual work. Choose GPT Image 2.5 when your task centers on generating or editing still images.
Choose a Shotyard plan for GPT Image 2.6. Generation uses credits, with the cost shown for your selected settings before you generate.
Learn about GIFs, SVGs, editing controls and the differences between model versions.
Use the GPT Image 2.6 AI Image Generator on Shotyard. Enter your prompt, choose your settings and start creating.
GPT Image 2.6 is OpenAI’s image generation and editing family, released as Images 2.6 on November 4, 2026. It includes Flare and Sunburst, and adds animated GIF creation and SVG output.
Yes. The release introduces GIF creation for ideas such as reactions, moving characters and visual loops. OpenAI also demonstrates turning a simple sketch into an animated character.
Yes. OpenAI announced SVG output for designs used across websites, apps and print. This does not establish direct structural editing of an uploaded SVG document.
Flare emphasizes faster iteration and improved layouts. Sunburst focuses on detailed compositions and stronger consistency across edits, particularly at medium and high quality.
Quality controls visual detail and fidelity. Complexity controls the planning devoted to meeting your request, with lower settings suited to focused edits and higher settings useful for more involved scenes.
The reviewed model documentation lists video and audio as unsupported. Animated GIF support does not imply MP4 generation, audio tracks or general image-to-video workflows.
The reviewed 2.6-specific documentation does not conclusively establish the maximum raster resolution. Resolution options documented for older models should not be treated as confirmed GPT Image 2.6 limits.
GPT Image 2.6 uses Shotyard credits. The generator displays the cost for your selected settings before you submit, so you can review it before each generation.
Name the subject or region to change, describe the desired result and identify the details you want to preserve. For a complex composition, specify the layout, lighting, style and exact wording separately.