GPT Image 2.6 Flare API
gpt-image-2.6-flaregpt-image-2.6-flare is the default tier of OpenAI's GPT Image 2.6: fast, high-quality generation with natural lighting, rich textures and complex layouts, transparent backgrounds included. It is the tier to build a pipeline on — in our test calls a 1K image came back in about 6 seconds with reasoning effort set to low, and about 11 seconds at the default high. 2.6 is a new set of abilities, not a sharper 2.5. It adds an output type and controls the previous generation did not have: animated GIF (2 to 64 frames, 20 to 2000 ms per frame), edits guided by a markup image — a transparent PNG with circles, arrows or notes drawn over the first reference — a mask where white marks the area to change and black the area to keep, up to 10 images per call, and reasoning_effort from low to max, which trades latency for detail. Sizes: pass aspect_ratio with 1K, 2K or 4K; an exact WIDTHxHEIGHT (both multiples of 16, longest edge up to 3840, ratio up to 3:1, 0.66 to 8.29 megapixels); a preset name; or auto. Quality is low, medium, high or auto, and on apimodels.app it defaults to medium — the model's own default is high, which cost about five times as much in our 1K test. Pricing is per token, the same way the model meters it: image output $33 per 1M tokens, image input $8.80 (cached $2.20), text input $5.50 (cached $1.375) and text output $11, each request rounded up to $0.0001. Quality, size, GIF, every reference image and the prompt all change the token count, so there is no fixed price per image. For a sense of scale, our measured requests on 5 November 2026 came to $0.0154 for 1K low, $0.028 for 1K medium, $0.031 for 2K low and $0.057 for 2K medium (16:9, one image). An upper bound is held when the request is accepted and the unused part is released when the image is ready. Know the limits before you build: SVG output is listed in the model's documentation but currently fails upstream, so apimodels.app returns a 400 for it for now; it will be enabled once it works. A markup image cannot be combined with a mask. For GIF, all frames are laid out on one sheet whose longest edge can be at most 3840 px: pass aspect_ratio with 1K / 2K / 4K (or no size) and apimodels.app shrinks the frames automatically so the sheet fits; with an exact size, keep it small enough yourself — 960x960 works for up to 16 frames. In our tests of the previous generation, masks behaved as hints rather than hard boundaries, and we have not measured 2.6 yet — for a strict local edit, also name the region in the prompt. Like any generative model it does not succeed every time; moderation refusals and upstream errors happen and are never charged. Results are kept for 30 days.
Generate Image
Billed per token: Image output $33 / 1M tokens · image input $8.80 (cached $2.20) · text input $5.50 (cached $1.375) · text output $11. Higher quality, larger sizes, GIF and reference images use more tokens. An upper limit is held when you submit and the unused part is released when the image is ready.
How much the model plans before drawing. Higher settings add detail and take longer.
Result
Generated image will appear here
Enter a prompt and click Generate
Billed per token
Image output $33 per 1M tokens, image input $8.80 (cached $2.20), text input $5.50 (cached $1.375), text output $11 — rounded up to $0.0001 per request. GIF, size, quality and references only change the token count. Measured requests: 1K low $0.0154, 2K medium $0.057. Failed requests are free.
Animated GIF, new in 2.6
Ask for output_format gif and get an animation — 2 to 64 frames, 20 to 2000 ms each. All frames are laid out on one sheet of at most 3840 px on the long edge; pass aspect_ratio with 1K / 2K / 4K and apimodels.app shrinks the frames automatically so it fits.
Point at what to change
Draw circles, arrows or notes on a transparent PNG and pass it as the markup image; the model edits the first reference accordingly. A mask (white = change, black = keep) is the alternative — not both at once. We have not measured mask strictness on 2.6, so name the region in the prompt too.
The fast default tier
About 6 seconds for a 1K image at reasoning effort low and about 11 seconds at high in our tests. Choose Flare when turnaround and volume matter; choose Sunburst when someone is going to zoom in.
About GPT Image 2.6 Flare
GPT Image 2.6 Flare is an Image Generation API provided by OpenAI. gpt-image-2.6-flare is the default tier of OpenAI's GPT Image 2.6: fast, high-quality generation with natural lighting, rich textures and complex layouts, transparent backgrounds included. It is the tier to build a pipeline on — in our test calls a 1K image came back in about 6 seconds with reasoning effort set to low, and about 11 seconds at the default high. 2.6 is a new set of abilities, not a sharper 2.5. It adds an output type and controls the previous generation did not have: animated GIF (2 to 64 frames, 20 to 2000 ms per frame), edits guided by a markup image — a transparent PNG with circles, arrows or notes drawn over the first reference — a mask where white marks the area to change and black the area to keep, up to 10 images per call, and reasoning_effort from low to max, which trades latency for detail. Sizes: pass aspect_ratio with 1K, 2K or 4K; an exact WIDTHxHEIGHT (both multiples of 16, longest edge up to 3840, ratio up to 3:1, 0.66 to 8.29 megapixels); a preset name; or auto. Quality is low, medium, high or auto, and on apimodels.app it defaults to medium — the model's own default is high, which cost about five times as much in our 1K test. Pricing is per token, the same way the model meters it: image output $33 per 1M tokens, image input $8.80 (cached $2.20), text input $5.50 (cached $1.375) and text output $11, each request rounded up to $0.0001. Quality, size, GIF, every reference image and the prompt all change the token count, so there is no fixed price per image. For a sense of scale, our measured requests on 5 November 2026 came to $0.0154 for 1K low, $0.028 for 1K medium, $0.031 for 2K low and $0.057 for 2K medium (16:9, one image). An upper bound is held when the request is accepted and the unused part is released when the image is ready. Know the limits before you build: SVG output is listed in the model's documentation but currently fails upstream, so apimodels.app returns a 400 for it for now; it will be enabled once it works. A markup image cannot be combined with a mask. For GIF, all frames are laid out on one sheet whose longest edge can be at most 3840 px: pass aspect_ratio with 1K / 2K / 4K (or no size) and apimodels.app shrinks the frames automatically so the sheet fits; with an exact size, keep it small enough yourself — 960x960 works for up to 16 frames. In our tests of the previous generation, masks behaved as hints rather than hard boundaries, and we have not measured 2.6 yet — for a strict local edit, also name the region in the prompt. Like any generative model it does not succeed every time; moderation refusals and upstream errors happen and are never charged. Results are kept for 30 days. Through APIMODELS platform, you can access this model via a unified API with transparent pay-as-you-go pricing. Current pricing: Image output: $$33.00 / 1M tokens, Image input: $$8.80 / 1M tokens, Image input (cached): $$2.20 / 1M tokens, Text input: $$5.50 / 1M tokens, Text input (cached): $$1.375 / 1M tokens, Text output: $$11.00 / 1M tokens.
Key Features
- Billed per token -- Image output $33 per 1M tokens, image input $8.80 (cached $2.20), text input $5.50 (cached $1.375), text output $11 — rounded up to $0.0001 per request. GIF, size, quality and references only change the token count. Measured requests: 1K low $0.0154, 2K medium $0.057. Failed requests are free.
- Animated GIF, new in 2.6 -- Ask for output_format gif and get an animation — 2 to 64 frames, 20 to 2000 ms each. All frames are laid out on one sheet of at most 3840 px on the long edge; pass aspect_ratio with 1K / 2K / 4K and apimodels.app shrinks the frames automatically so it fits.
- Point at what to change -- Draw circles, arrows or notes on a transparent PNG and pass it as the markup image; the model edits the first reference accordingly. A mask (white = change, black = keep) is the alternative — not both at once. We have not measured mask strictness on 2.6, so name the region in the prompt too.
- The fast default tier -- About 6 seconds for a 1K image at reasoning effort low and about 11 seconds at high in our tests. Choose Flare when turnaround and volume matter; choose Sunburst when someone is going to zoom in.
Examples by scenario
Use Cases
E-commerce Product Images
Generate high-quality product visuals for online stores, ads, and marketing materials.
Social Media Content
Create eye-catching visual content for social platforms to boost engagement and brand visibility.
Game Asset Generation
Produce concept art for characters, scenes, and props to accelerate game development.
Marketing Materials
Design posters, banners, and promotional graphics at a fraction of traditional design costs.
Why APIMODELS
- Unified API -- One API key to access all models, no need to register on multiple platforms
- Transparent Pricing -- Pay-as-you-go with per-call pricing published on every model page, ideal for indie developers and startups
- Instant Access -- Start using immediately after signup, supports Stripe and Alipay payments
- Full Documentation -- Detailed API docs with code examples in cURL, Python, and Node.js
Frequently Asked Questions
How much does GPT Image 2.6 Flare cost?
GPT Image 2.6 Flare is available through APIMODELS at: Image output: $$33.00 / 1M tokens, Image input: $$8.80 / 1M tokens, Image input (cached): $$2.20 / 1M tokens, Text input: $$5.50 / 1M tokens, Text input (cached): $$1.375 / 1M tokens, Text output: $$11.00 / 1M tokens. Billing is pay-as-you-go — you only pay for what you generate.
How to use GPT Image 2.6 Flare API?
Sign up at APIMODELS, get your API key, and call our unified API endpoint. We provide detailed API documentation with code examples in cURL, Python, and Node.js.
What is the difference between APIMODELS and the official OpenAI API?
APIMODELS offers the same GPT Image 2.6 Flare model through our aggregation platform. We provide a unified API interface so you do not need separate accounts for each provider - one API key to access all models.
How much does gpt-image-2.6-flare cost?
It is billed per token, like the model itself: image output $33 per 1M tokens, image input $8.80 (cached $2.20), text input $5.50 (cached $1.375) and text output $11, each request rounded up to $0.0001. There is no fixed per-image price because quality, size, format (GIF), reference images and prompt length all change the token count. As a guide, our measured requests on 5 November 2026 cost $0.0154 for 1K low, $0.028 for 1K medium, $0.031 for 2K low and $0.057 for 2K medium. An upper bound is held when you submit and the unused part is released; failed requests are not charged.
What is new in GPT Image 2.6 compared with GPT Image 2.5?
A new output type and new controls: animated GIF (2–64 frames), edits guided by a markup image, a mask where white marks the area to change, up to 10 images per call, and reasoning_effort from low to max. Sizes can be exact pixel dimensions up to 3840 on the long edge. GPT Image 2.5 stays available on apimodels.app with its fixed per-image prices, so pick 2.5 when you want a predictable price and 2.6 when you need these abilities.
Should I use Flare or Sunburst?
Both take the same parameters and bill from the same token price list. Flare is the default, faster tier — about 6 seconds for 1K at reasoning effort low in our tests — and suits social posts, catalogue images and high-volume pipelines. Sunburst spends more time per image for finer detail and edits held tightly to the instruction; use it for premium visuals that will be inspected up close.
Can gpt-image-2.6-flare make animated GIFs?
Yes. Set output_format to gif and optionally frame_count (2–64, default 9) and frame_duration_ms (20–2000, default 200). All frames are laid out on one sheet whose longest edge can be at most 3840 px: with aspect_ratio plus 1K / 2K / 4K (or no size) apimodels.app shrinks the frame size automatically so the sheet fits; with an exact size, keep it small enough — 960x960 works for up to 16 frames. SVG output is currently unavailable: it fails upstream, so apimodels.app returns a 400 for it until it works.
How does GPT Image 2.6 Flare compare to other Image Generation models?
On APIMODELS, GPT Image 2.6 Flare runs alongside 60+ models on one API key and one balance, so choosing is about fit, not lock-in. It supports Text to Image & Edit, Animated GIF output, Markup-image and mask edits, Reasoning effort low to max, Per-token pricing, and you can weigh it on price and capability against other Image Generation models, then switch by changing a single model-name string — no new account or integration. Browse every Image Generation option with live pricing at apimodels.app/models.
What can GPT Image 2.6 Flare do?
GPT Image 2.6 Flare supports: Text to Image & Edit, Animated GIF output, Markup-image and mask edits, Reasoning effort low to max, Per-token pricing. See the APIMODELS docs for full parameters and call examples.
Can I access the GPT Image 2.6 Flare API from anywhere (incl. China)?
Yes. APIMODELS exposes GPT Image 2.6 Flare through a single unified API and one key — no separate provider accounts, and no need to handle each provider's regional network access yourself.
What payment methods are supported?
We support Stripe (Visa, Mastercard, and other international cards) and Alipay. Credits are available instantly after payment.