Sign InOpen Brain
VercelEngineering PostOfficial Source

GPT Image 2.5 Flare and Sunburst now available on AI Gateway

Vercel AI Gateway adds GPT Image 2.5 Flare for faster iteration and Sunburst for tighter composition control. Both support generation, reference-based edits, and transparent backgrounds.

Vercel · Sep 8, 2026
Open Source Open MarkdownOpen JSON
Source Summary

AI Gateway now exposes **openai/gpt-image-2.5-flare** and **openai/gpt-image-2.5-sunburst**. Both generate and edit images, accept references, support detailed layouts and transparent backgrounds, and can preserve surrounding elements during targeted edits.

Practical Implication

Use **Flare for faster iteration** and switch to **Sunburst for tighter control** over composition and detail. The same gateway also offers usage and cost tracking, retries, failover, budgets, and routing rules.

Agent-Ready Context
AI Gateway now exposes **openai/gpt-image-2.5-flare** and **openai/gpt-image-2.5-sunburst**. Both generate and edit images, accept references, support detailed layouts and transparent backgrounds, and can preserve surrounding elements during targeted edits.

Use **Flare for faster iteration** and switch to **Sunburst for tighter control** over composition and detail. The same gateway also offers usage and cost tracking, retries, failover, budgets, and routing rules.

Sunburst takes longer to generate, while the material provides no latency, quality, or cost comparison between the models. Provider pricing is passed through without inference markup or a platform fee, including BYOK requests.
Connected Context · Feed7 Judgment

This adds an explicit speed-versus-control routing choice within one image-model family, making Gateway selection more actionable than candidates that merely establish another available route. It also reinforces a common generation, reference, and targeted-edit interface, but provides no evidence that either variant outperforms existing Gateway image models or quantifies the tradeoff.

Context Map
infraimage#gateways#generative-media#model-selection
Uncertainty
Sunburst takes longer to generate, while the material provides no latency, quality, or cost comparison between the models. Provider pricing is passed through without inference markup or a platform fee, including BYOK requests.