$ cat wiki/models/gpt-image-2-6-flare.md
GPT-Image-2.6 Flare
modelupdated 2026-11-05created 2026-11-05
Spec
| Attribute | Value |
|---|---|
| Developer | OpenAI |
| Released | 2026-11-04 |
| Announced | 2026-11-04 |
| Context window | unknown |
| Pricing | Text input $5/M, cached $1.25/M; image input $8/M, cached $2/M, output $30/M |
| License | unknown |
| Availability | OpenAI API: v1/images/generations, v1/images/edits |
| Release and developer: (source) Token prices and supported endpoints: (source) The reviewed sources do not specify a context window or license grant. |
Release Date
Announced and available in the API on 2026-11-04. The broader Images 2.6 experience also launched in ChatGPT and Codex; this does not establish a user-selectable Flare variant in those products. (source)
Benchmarks
OpenAI positions Flare for fast interactive iteration and says low-quality generation can take as little as six seconds. That is a vendor claim, not an independently reproduced latency distribution. (source)
Use Cases
- Generate and edit images from text and image inputs with
gpt-image-2.6-flare. (source) - Use image markup for targeted edits, SVG export and animated GIF output. The complexity setting controls planning separately from image quality; output quality and workflow suitability require evaluation on the intended task. (source) (source)
Compared To
GPT-Image-2.6 Sunburst is the other Images 2.6 API option. OpenAI distinguishes fast iteration from detailed visual work; the reviewed model-reference pages list the same token rates. These sources do not establish equivalent output quality or equal per-image cost. (source) (source)
Sources
- Announcement — 2026-11-04; selected paragraphs captured locally. (source)
- Model reference — read 2026-11-05; complete Markdown response captured locally. (source)