Introducing GPT-image-2.6 Flare and Sunburst in Microsoft Foundry
OpenAI's GPT-image-2.6 is rolling out in Microsoft Foundry as two models on Global Standard: Flare, built for speed (as little as six seconds per image, per OpenAI), and Sunburst, built for quality and consistency across edits. New in this generation are markup-based editing of specific image areas, complexity control, and GIF and SVG output. Both models are priced the same per million tokens.
- status
- Generally available
SVG output means icons and illustrations can go straight into a design system without being redrawn, and markup editing cuts down on rerunning a whole prompt to fix one detail. Since both models cost the same per token, the choice between them comes down to speed versus quality. Prototype with Sunburst, move high-volume drafts to Flare, and compare the results on your own prompts. Check the region list before you plan production, since Global Standard processing may matter for data residency.
Official sources
- techcommunity.microsoft.comIntroducing GPT-image-2.6 Flare and Sunburst in Microsoft Foundry