GPT Image 2.6 Flare API
Generate GPT Image 2.6 images from your own code, in both versions: Flare (fast) and Sunburst (detail). Pick an output format of png, jpeg, webp, an animated gif or a scalable svg, five quality tiers from low to max, and sizes up to 4K, with optional input images for edits. Several providers resell this model at very different prices; every request is priced across them and runs on the cheapest one that supports your settings.
67% less than the priciest provider (7 compared). Same model, same output.
As of , GPT Image 2.6 Flare costs from $0.016 per image on Scenetra, billed at provider cost with no markup. Each run is priced across 7 providers and runs on the cheapest one that supports your settings.
- Low - 1K
- $0.016
- Low - 2K
- $0.02
- Low - 4K
- $0.042
- Medium - 1K
- $0.028
- Medium - 2K
- $0.034
- Medium - 4K
- $0.073
- High - 1K
- $0.032
- High - 2K
- $0.038
- High - 4K
- $0.085
- Extra High - 1K
- $0.055
- Extra High - 2K
- $0.065
- Extra High - 4K
- $0.147
- Max - 1K
- $0.119
- Max - 2K
- $0.14
- Max - 4K
- $0.325
- GIF, 8 frames (Medium, 1:1)
- $0.03
- GIF, 16 frames (Medium, 1:1)
- $0.045
- SVG - Medium
- $0.049
- SVG - High
- $0.057
Quickstart
Async by design: submit a generation, get a run id back instantly, poll until it completes (or pass a callbackUrl).
1 — Start a generation
curl -X POST https://app.scenetra.com/api/v1/generate/gpt-image-26-flare \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{"prompt":"a watercolor fox in falling snow, soft light","quality":"medium","aspect_ratio":"1:1"}'Response (202)
{
"runId": "run_abc123",
"status": "PENDING",
"model": "gpt-image-26-flare",
"estimatedCredits": 30,
"statusUrl": "/api/v1/runs/run_abc123"
}2 — Poll until COMPLETED (every 2-5 seconds)
curl https://app.scenetra.com/api/v1/runs/run_abc123 \
-H "Authorization: Bearer YOUR_API_KEY"Response
{
"runId": "run_abc123",
"status": "COMPLETED",
"progress": { "completed": 1, "total": 1 },
"outputs": [
{ "type": "image", "url": "https://images.scenetra.com/..." }
],
"error": null,
"estimatedCredits": 30
}Request fields
| Field | Type | Description | Default |
|---|---|---|---|
| prompt* | string | Describe the image to generate, or the edit to apply to the input images. | — |
| image_urls | string[] (https URLs) | Optional images to edit or use as reference (up to 16). Leave empty for pure text-to-image. | — |
| resolution | string | Output resolution tier.1K · 2K · 4K | 1K |
| aspect_ratio | string | Aspect ratio of the generated image. 'auto' lets the model choose.auto · 21:9 · 16:9 · 3:2 · 4:3 · 5:4 · 1:1 · 4:5 · 3:4 · 2:3 · 9:16 | auto |
| quality | string | Rendering quality. Low is a fast draft; xhigh and max push detail for final assets. The price difference is large.low · medium · high · xhigh · max | medium |
| background | string | Transparent requires PNG, WebP or SVG output. Auto lets the model decide.auto · transparent · opaque | auto |
| num_images | number | How many images to generate (1-4). Billed per image. | 1 |
| output_format | string | File type of the result. GIF returns a short looping animation; SVG returns an editable vector file.png · jpeg · webp · gif · svg | png |
| complexity | string | How much the model plans before rendering. Higher helps dense text and complex layouts; no extra cost.low · medium · high | high |
| frame_count | number | GIF only: number of frames (2-8, or 16). | 9 |
| frame_duration_ms | number | GIF only: how long each frame shows, in milliseconds (20-2000). | 200 |
| callbackUrl | string (https URL) | Optional. When the run finishes (completed or failed), we POST the final run status to this URL — same shape as the poll response, plus model, params, and chargedCredits. Best-effort; polling remains the source of truth. | |
Unknown fields are rejected with a helpful 400. Media fields take https URLs. There is no provider parameter — every run is routed to the cheapest capable provider automatically.
Good to know
Billing
Runs bill your account credits at provider cost with 0% markup. The submit response includes the credit estimate; failed generations are refunded automatically. Add ?dryRun=1 to get the estimate without running anything.
Errors
Errors are always { "error": { "code", "message" } } with stable codes: unauthorized, model_not_available, invalid_parameter, insufficient_credits, rate_limited. A failed run is not an HTTP error — poll responses report status: "FAILED" with the reason.
GPT Image 2.6 Flare API FAQ
How do I get a GIF or SVG?+
Set output_format to gif or svg. GIFs are animated from a frame sheet; SVGs are vector traces you can scale to any size. Both cost about the same as a regular image at the same quality.
What is the difference between Flare and Sunburst?+
Flare (gpt-image-26-flare) is tuned for speed and Sunburst (gpt-image-26-sunburst) for detail. Every provider prices them the same, and both take the same fields.
Is this the official OpenAI API?+
No. Scenetra routes each call to a provider that serves GPT Image 2.6 (the same model and output) and bills you in Scenetra credits from one balance.
How much does the GPT Image 2.6 Flare API cost?+
As of November 7, 2026, GPT Image 2.6 Flare costs from $0.016 per image on Scenetra, billed at provider cost with no markup. Each run is priced across 7 providers and runs on the cheapest one that supports your settings. Rates: Low - 1K $0.016; Low - 2K $0.02; Low - 4K $0.042; Medium - 1K $0.028; Medium - 2K $0.034; Medium - 4K $0.073; High - 1K $0.032; High - 2K $0.038; High - 4K $0.085; Extra High - 1K $0.055; Extra High - 2K $0.065; Extra High - 4K $0.147; Max - 1K $0.119; Max - 2K $0.14; Max - 4K $0.325; GIF, 8 frames (Medium, 1:1) $0.03; GIF, 16 frames (Medium, 1:1) $0.045; SVG - Medium $0.049; SVG - High $0.057. With default settings (1K), a run is estimated at about 30 credits. The submit response returns the estimate for your exact settings, and adding ?dryRun=1 prices a request without running it.
Which provider runs my request?+
The cheapest one that supports your settings, chosen per request. There is no provider parameter: the same GPT Image 2.6 Flare request is priced across every provider we route to and runs on the lowest quote.
Do failed generations cost credits?+
No. Credits are reserved when you submit and refunded automatically if the run fails. A failed run is not an HTTP error: polling returns status FAILED with the reason.
Do I have to poll, or can I get a webhook?+
Both work. Pass callbackUrl and we POST the final run status to it when the run completes or fails. Delivery is best-effort, so polling GET /api/v1/runs/{runId} remains the source of truth.