OpenAI’s API changelog lists two new image models released on November 4, 2026: GPT-Image-2.6 Sunburst (gpt-image-2.6-sunburst) and GPT-Image-2.6 Flare (gpt-image-2.6-flare).
What to know
- Two models, two jobs: OpenAI’s image generation guide describes Sunburst as for detailed visual work and Flare as for fast iteration. Both generate and edit images through the Image API (
v1/images/generationsandv1/images/edits). - Complexity control: a
complexitysetting (low,mediumorhigh) sets how much the model plans before it renders. OpenAI says both models default tohigh, that lower settings suit focused edits such as a background change, and that higher settings help with dense text and complex layouts. It also warns that higher settings can add latency and unwanted changes, and says adjusting complexity has no separate fee. - Markup-guided editing: you send a transparent PNG annotation layer (
markup) with an edit request to show where the change should happen. OpenAI says it must match the first input image’s size, and cannot be combined withmaskorprompt_rewriter. The changelog says markup-guided editing was also added for supported GPT Image 2.5 and 2.6 models. - New output types: OpenAI says both models support SVG and animated GIF output.
- Pricing: the model pages list, per million tokens, $5 for text input, $8 for image input and $30 for image output. Check OpenAI’s pages for current rates and rate limits.
- Where: the pages describe these as API image models only. We did not see a ChatGPT rollout stated.
Why it matters
Complexity control gives you a speed-versus-care dial for image work, and markup-guided editing lets you point at the area to change instead of describing it. We wrote our own tips in how to prompt GPT-Image-2.6. See also GPT Image, Midjourney vs GPT Image and the model tracker.
Source: OpenAI API changelog. We summarise in our own words; check the source for the full details.