GPT Image 2.6Flare and Sunburst

GPT Image 2.6 is OpenAI's image generation and editing model, released in November 2026. It adds animated GIF and SVG output, markup-guided editing and a complexity setting on top of what GPT Image 2.5 does. Two variants are available to developers: GPT-Image-2.6 Flare and GPT-Image-2.6 Sunburst.

Yes. Sign up at faktry.ai and you get 50 free credits instantly, with no credit card required. A standard high-quality GPT Image 2.6 image costs 10 credits and a low-quality draft costs 6, so the free credits go a long way. After that there is no subscription, you pay only for what you generate.

Flare is built for fast iteration and suits drafts, social content, GIFs and high-volume work. Sunburst is built for detailed visual work, such as campaign creative, polished product imagery and precision edits. They cost the same per image on faktry, so start with Flare and switch to Sunburst when the result needs more detail.

GPT Image 2.6 can make animated GIFs and SVG vector files from a prompt, which GPT Image 2.5 cannot. On faktry it also supports a markup layer for pointing at the area an edit should change, and a complexity setting for how much planning the model does. Both generations offer the same five quality tiers and take up to 16 reference images per edit.

Choose GPT Image 2.6 in Generate Image, set the output format to GIF, then pick the number of frames (2 to 16) and how long each frame shows (20 to 2000 milliseconds). Describe the motion in your prompt, for example a character reacting to something in a short loop. GIFs are generated from text, they cannot be made by editing an existing image.

Yes. Set the output format to SVG in Generate Image. You get one static vector file per request, ready for websites, apps and print. SVGs contain no scripts, external resources or embedded photos, and they are generated from text, not from an edit.

Markup lets you show the model where a change should happen. In Edit Image, upload a transparent PNG drawn over your image as the markup layer. It must have exactly the same pixel size as the first image you upload. The model then makes the change at the marked spot. The markup layer does not count as a reference image.

Complexity sets how much planning the model does before it renders, separately from image quality. OpenAI recommends low complexity for quick edits, and medium or high for scenes with many elements or constraints. If you leave it unset, OpenAI uses high. It carries no extra fee.

Use high for finished work, it is the default. Drop to low or medium for drafts, thumbnails and layout tests, where a render costs 6 or 9 credits instead of 10. Reach for xhigh or max only when the extra detail is genuinely the point, since they cost 20 and 40 credits.

Up to 16 per edit. Editing starts at 8 credits plus 2 per reference image, so pass the references you actually need rather than everything to hand.

Yes. Images you generate on faktry can be used commercially, including marketing, advertising, client work and products.