GPT Image 2.5 — Flare and Sunburst

GPT Image 2.5 has two distinct model IDs. There is no gpt-image-2.5 model or alias.

The official snapshots are gpt-image-2.5-flare-2026-09-08 and gpt-image-2.5-sunburst-2026-09-08. Use the undated IDs above with Quotaflow. Catalog listing does not establish serving availability; check your authenticated /models response. Both IDs are serving as of 2026-09-16, on /images/generations and on /images/edits including masked edits.

Endpoints and request parameters

Use the synchronous Images API at https://api.quotaflow.ai/openai/v1/images/generations or /openai/v1/images/edits. These models do not support the Responses API.

• quality: low, medium, high, xhigh, max, or auto; defaults to auto. The supplied value is forwarded unchanged; unknown values return HTTP 400 with param: "quality".
• size: 1024x1024, 1536x1024, 1024x1536, or a custom WIDTHxHEIGHT. Both edges must be multiples of 16, neither edge may exceed 3840, aspect ratio must be between 1:3 and 3:1, and total pixels must be between 655,360 and 8,294,400.
• n requests multiple output images; edits also accept a mask.
• output_format: png, jpeg, or webp. background: "transparent" requires png or webp. output_compression and moderation (auto or low) are forwarded when supplied.
• OpenAI documents partial_images from 0 to 3. Quotaflow currently rejects streaming and partial_images; use the synchronous endpoint.
• OpenAI does not document input_fidelity for 2.5. Quotaflow applies the same handling as GPT Image 2 and does not forward it.

Omitting model still selects gpt-image-2, whose omitted quality remains medium and whose size: "auto" becomes 1024x1024.

curl https://api.quotaflow.ai/openai/v1/images/generations \
  -H "Authorization: Bearer $QUOTAFLOW_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gpt-image-2.5-flare","prompt":"A clean editorial illustration of an AI workflow.","size":"1024x1024","quality":"xhigh","response_format":"b64_json"}'

Standard token prices

Both models use the same standard rates. Prices are USD per 1 million tokens; Batch is not offered.

Settlement uses reported token usage. GPT Image 2.5 does not use GPT Image 2 token estimates: OpenAI states that its GPT Image 2 calculator does not estimate GPT Image 2.5 token consumption. A temporary balance hold is separate from the final charge.

Official sources, read September 13, 2026: pricing, Flare, Sunburst, and image generation guide.