OpenAI released Images 2.6 on November 4, 2026, with a wider set of outputs for people making apps, games and visual content. The headline addition is animated GIF creation, but the developer changes are also concrete: markup-guided edits, SVG export and a new complexity control in the API. OpenAI offers two API models, GPT-Image-2.6 Flare for quicker iteration and Sunburst for more detailed visual work.
What can Images 2.6 make and edit?
The release supports generating and editing images, then turning ideas such as a character reaction or a simple storyboard into a GIF. OpenAI's examples include animated characters for a product interface and visual content for social sharing. An app maker could use that kind of output as an early concept or an asset to refine. The examples demonstrate possible workflows; they do not show that every prompt will produce a production-ready animation.
For developers, image markup lets an API edit request indicate a specific part of an image to change. SVG export provides a vector output option for designs that need to scale across an interface or print. The Image API guide says both Flare and Sunburst support markup-guided edits, SVG output, animated GIF output and a complexity setting. That makes the update relevant to software products that let users create or revise visuals, not only to standalone image generation.
How do Flare, Sunburst and complexity differ?
OpenAI positions Flare around speed and Sunburst around higher-detail results and consistency across edits. It says Images 2.6 generates images 50% faster than its previous generation, and that Flare at low quality can produce an image in as little as six seconds. Those are OpenAI's release claims, not timings measured by VIBPS. Real latency will depend on the request, output settings and service conditions.
The API's complexity setting controls how much planning the model applies to a request. OpenAI documents low, medium and high values, with high as the default for both models. Quality is a separate control for visual fidelity. A quick, focused edit may suit lower complexity, while a dense scene can need more planning. Neither setting guarantees a desired result, so a creator still needs to inspect the generated output and its rights and safety context before publishing it.
Where is Images 2.6 available now?
OpenAI says Images 2.6 is available globally in ChatGPT and Codex across subscription tiers on desktop, mobile and web. It also says GPT-Image-2.6 Flare and Sunburst are available through the API, where usage has separate pricing. That is an availability statement from OpenAI as of the announcement, not a claim that every third-party product has integrated the models. Developers should check the current API guide and their own account access before choosing an implementation.
For a maker, a faster visual workflow can help explain an app or test a character idea, but the visual still needs to represent the real product honestly. VIBPS is a social network for software: makers can publish apps they own and connect the work to their profile and audience. This article is based on OpenAI's release and documentation; VIBPS has not benchmarked or generated images with 2.6 for this report.