OpenAI Image API Update: Flare Generates Images in 6 Seconds, With Built-in GIF and SVG Export Support
OpenAI has officially released the Images 2.6 API, as well as the GPT-Image-2.6 Flare and Sunburst API models. Flare can generate images in as fast as 6 seconds. Both new models add GIF output, SVG export, markup editing and custom complexity parameters. This post covers the key information regular users and developers should pay attention to respectively.
OpenAI has updated its entire image API product line, launching Images 2.6 alongside two new API models under the GPT-Image-2.6 family: Flare and Sunburst. According to MarkTechPost's report published on November 4, 2026, Flare can produce finished images in as little as 6 seconds; both models now come with GIF output support, SVG export, markup editing features, and adjustable complexity parameters.
First, on generation speed: Flare has cut the end-to-end generation time to a minimum of 6 seconds. For regular end users, this means much shorter waiting time for their custom images. For product developers, this 6-second turnaround means image generation can be integrated far more naturally into interactive workflows, such as in-place editing, real-time preview, and bulk iterative testing.
Next, on new output formats: GIF is ideal for short dynamic content, while SVG is far more useful for icons, UI assets, and any graphics that require lossless scaling. The fact that both models support these two output formats proves this update is not only focused on improving static single-image generation, but also completes the export stage in full real-world creative workflows.
Markup editing and the complexity parameters are also highly valuable for developers. The former can be understood as targeted markup-style modification on generated images, while the latter allows API callers to control the level of detail and complexity of the final output. The exact feature behavior, valid parameter ranges and practical performance still need to be verified with official API documents and real-world testing. Only the full feature list is confirmed at present, while details on pricing, usage quotas, and quality differences across the two models are not covered in the original announcement.
Source: MarkTechPost Original Article, by Asif Razzaq.
发布时间: 2026-11-05 15:03