GPT Image 2.6 — GIFs, SVGs and faster edits
GPT Image 2.6 generates images, animated GIFs and SVGs, with complexity control and reference-image editing. Choose a configured model below to try it on SeedRouter.
- Text to image
- Image editing
- Animated GIF
- SVG output
- Complexity control
Try GPT Image 2.6
Generate images, animated GIFs or SVGs with the same parameters available in the API.
gpt-image-2.6-sunburst-official
Generate from a prompt or edit reference images. Submit once; results appear when the task completes.
required · up to 32,000 characters
Reference images for image-to-image and editing. Files are uploaded and sent as URLs; on Official IDs each one adds image input tokens.
Let the model choose the image dimensions.
auto or WIDTHxHEIGHT · custom sizes supported
Reference images for image-to-image and editing. Files are uploaded and sent as URLs; on Official IDs each one adds image input tokens.
Transparent PNG, at most 50 MB, matching the first reference image dimensions. Transparent areas are edited.
auto, low, medium, high, xhigh or max. Default: auto.
1 to 10 · default 1
auto, opaque or transparent. JPEG cannot be transparent.
default png
Choose JPEG to set compression.
default auto
Greyed-out options relax content filtering. They are available through the API only, not in the playground.
Planning effort: low, medium or high. Default: high.
Available with GIF output.
Available with GIF output.
Available with SVG output; one output per request.
Available with SVG output; one output per request.
Reference images for image-to-image and editing. Files are uploaded and sent as URLs; on Official IDs each one adds image input tokens.
Transparent PNG, at most 50 MB and 16,777,216 pixels. Must match the first reference image after orientation. Cannot be combined with a mask.
Your stable end-user identifier, at most 128 characters. Never filled automatically.
- prompt is required
- prompt: Invalid input: expected string, received undefined
No measured estimate for these settings. Actual cost depends on image size, quality, prompt and reference images.
GPT Image 2.6 samples
Three GPT Image 2.6 requests from our tests on November 5, 2026, one per output format. The GIF and SVG are the files the API returned; the PNG is shown as WebP.
Still image
Sunburst · low · 1024×1024
Animated GIF
Flare · low · 4 frames × 200 ms · 512×512
SVG
Sunburst · low · vector paths
GPT Image 2.6 pricing
Current SeedRouter prices for the configured GPT Image 2.6 models. Token rates and per-image rates update from the pricing catalogue.
| Category | gpt-image-2.6-flare-officialPrice · 1M tokens | gpt-image-2.6-sunburst-officialPrice · 1M tokens |
|---|---|---|
| Text input | $3.75 | $3.75 |
| Image input | $6.00 | $6.00 |
| Cached text input | $0.9375 | $0.9375 |
| Cached image input | $1.50 | $1.50 |
| Image output | $22.50 | $22.50 |
| Per image | — | — |
What is GPT Image 2.6?
GPT Image 2.6 is OpenAI's image generation and editing model, launched November 4, 2026 in ChatGPT, Codex and the API. GPT Image 2.6 turns a prompt or an image into a still image, an animated GIF or an SVG, and OpenAI says it generates images 50% faster than before.
In the API, GPT Image 2.6 comes as two models. Flare can return a low-quality image in as little as six seconds, says OpenAI. Sunburst suits detailed compositions and keeps edits consistent at medium or high quality.
Official parameters reviewed November 5, 2026
Where GPT Image 2.6 runs today
Access to GPT Image 2.6 in OpenAI's products does not carry over to SeedRouter.
ChatGPT and Codex
GPT Image 2.6 is available worldwide in ChatGPT and Codex on every subscription tier, on web, desktop and mobile.
OpenAI API
Flare and Sunburst run on OpenAI's image generation and edit endpoints.
SeedRouter
Use the Playground or submit an asynchronous image task with your SeedRouter API key.
What's new in GPT Image 2.6
OpenAI's changelog lists three additions that arrived with GPT Image 2.6.
Animated GIFs
Describe a subject, its motion and the loop, and GPT Image 2.6 returns a finished GIF. Animate a mascot or a product shot without a video tool.
SVG output
Ask for SVG and get editable paths and colors instead of pixels, ready for a website, an app or print.
Complexity setting
Turn complexity down for a quick background swap and up for dense text or busy scenes. It adds no fee.
What to make with GPT Image 2.6
Ideas from OpenAI's GPT Image 2.6 launch examples, not SeedRouter test results.
Brand mascots that move
Describe a mascot and its movement to create a looping GIF for a campaign or game prototype.
Logos and icons
Generate an SVG mark you can recolor and resize, then refine it by hand.
Posters and product images
OpenAI says GPT Image 2.6 Flare lays out posters and product imagery better, which suits fast rounds of drafts.
Start using GPT Image 2.6
- 01
Pick Flare or Sunburst
Use Flare for many quick drafts. Use Sunburst when detail and consistent edits matter more than speed.
- 02
Choose the output
Still image, animated GIF or SVG. GPT Image 2.6 makes SVGs on generation only, so plan SVG edits as new prompts.
- 03
Set complexity
Start low for focused edits and raise it for crowded scenes.
- 04
Submit and retrieve your result
Submit once, keep the task ID and poll for the finished file URL. Download GIFs and SVGs in their original format.
GPT Image 2.6 API
Use one asynchronous image endpoint for generation and editing. The API reference lists every parameter and its constraints.
Generate or edit
POST /v1/images/generations accepts a prompt and optional reference image URLs. Poll the returned task ID for the result.
GIF and SVG controls
Set output_format to gif or svg. GIF supports frame count and duration; SVG supports optional author model and reasoning settings.
Current model prices
The price table above reads the current model rates. Token-priced requests settle using actual usage.
GPT Image 2.6 vs GPT Image 2.5
Both ship as Flare and Sunburst at the same token rates; GPT Image 2.6 adds new outputs and controls.
| Feature | GPT Image 2.6 | GPT Image 2.5 |
|---|---|---|
| Animated GIF output | Yes, 2–16 frames | No |
| SVG output | Yes, one per request | No |
| Complexity setting | Low, medium, high | No |
| Markup-guided edits | Yes | Yes |
| Token billing | Input and output tokens | Input and output tokens |
| On SeedRouter | Available now | Available now |
OpenAI's changelog lists complexity, SVG and GIF as new with GPT Image 2.6 and added markup editing to both releases. OpenAI API changelog, November 4, 2026
GPT Image 2.6 limits to know
SVG is generation only
GPT Image 2.6 does not return SVGs from the edit endpoint, and an SVG cannot be an input image.
One SVG per request
SVG output needs n set to 1, has no compression setting and holds vector shapes only.
GIF size budget
Frame count, frame size and the number of GIFs share one pixel budget.
Higher complexity is slower
OpenAI warns that high complexity can add latency and unrequested changes.
Why use GPT Image 2.6 on SeedRouter
One key for every model
Use one SeedRouter API key for GPT Image 2.6 and the other models available to your account.
Credit that never expires
Top up once and pay for what you generate. No subscription; unused credit stays on your account.
Failed requests are free
If a request ends without an image, nothing is charged.
Image models side by side
Compare GPT Image, Nano Banana, Seedream and more on one price list.
GPT Image models you can use now
GPT Image models live on SeedRouter today.
Create and edit images with GPT Image 2 on SeedRouter. Use text prompts, reference images and transparent backgrounds, with a live playground and usage-based pricing.
View pricingGenerate and edit images with GPT Image 2.5 Flare or Sunburst on SeedRouter. Six quality steps up to max, transparent backgrounds, a live playground and pay-as-you-go pricing.
View pricingGPT Image 2.6 guides
Articles about this model from the SeedRouter blog.
Common questions
What is GPT Image 2.6?+
GPT Image 2.6 is OpenAI's image generation and editing model, released November 4, 2026. It ships in the API as gpt-image-2.6-flare and gpt-image-2.6-sunburst, and in ChatGPT and Codex as Images 2.6. GPT Image 2.6 returns still images, GIFs and SVGs.
Can I use GPT Image 2.6 on SeedRouter?+
Yes. Select an available model in the Playground or use its ID from GET /v1/models. Current model prices appear on this page.
How much does GPT Image 2.6 cost?+
Current SeedRouter rates appear in the pricing table on this page. Token-priced generation depends on actual input and output usage; per-image models use the configured image price. Choose a model to see its billing basis.
GPT Image 2.6 Flare vs Sunburst: which should I use?+
Pick Flare for fast iteration: OpenAI says it can finish a low-quality image in as little as six seconds. Pick GPT Image 2.6 Sunburst for detailed work, where it gives higher quality and more consistent edits at medium or high quality.
Can GPT Image 2.6 make animated GIFs?+
Yes. Set the output format to GIF on a GPT Image 2.6 request, with 2 to 16 frames (default 9) and at least 20 ms per frame (default 200). A GIF cannot use a mask, and its frames must fit the model's size limits; size auto picks a frame size that fits.
Does GPT Image 2.6 export SVG files?+
Yes, from generation requests, one image at a time. GPT Image 2.6 renders a reference image, then writes static SVG code from it, so you pay for both steps.
What does complexity do in GPT Image 2.6?+
It sets how much GPT Image 2.6 plans before rendering: low, medium or high, with high as the default. OpenAI recommends low for most edits and medium or high for crowded scenes. Complexity adds no fee.
What changed from GPT Image 2.5 to GPT Image 2.6?+
GPT Image 2.6 adds GIF output, SVG output and the complexity setting, and OpenAI says it generates images 50% faster. Token rates and the low-to-max quality steps are unchanged; markup edits work on both.
Create with GPT Image 2.6
Try a prompt in the Playground, then use the same parameters in your application.