Built for workflows where precise image edits matter most.
Back to models
Editing precision Quality control Generation and inpainting
OpenAI
GPT Image 2.5 Sunburst
FeaturedMost capable image generation and editing20% offInputImageTextOutputImage
API formatOpenAI · Image Generation
OpenAI's most capable image generation and editing model, tuned for workflows where editing precision matters most.
gpt-image-2.5-sunburstText input$4.00$5.00per million tokens
Text output$24.00$30.00per million tokens
Cache read$1.60per million tokens
Image input$6.40$8.00per million tokens
Context window--Maximum context per request
Max output--
BillingPer token
HealthNo dataNo performance samples in the last hour
Model details
Capabilities, benchmarks, and typical use
GPT Image 2.5 Sunburst is OpenAI's most capable model for image generation and editing, with default snapshot gpt-image-2.5-sunburst-2026-09-08. It accepts text and image inputs and returns images, supports the low, medium, high, xhigh, max, and auto quality settings, and supports inpainting. It is available on the image generation and image edit endpoints and as the image generation tool of the Responses API. Text tokens are billed at $5.00 per 1M input and $1.25 per 1M cached input; image tokens are billed at $8.00 per 1M input, $2.00 per 1M cached input, and $30.00 per 1M output. Batch processing halves the image output rate to $15.00 per 1M tokens. OpenAI does not publish a context window for this image model, so the token limit fields remain null.
Supports low, medium, high, xhigh, max, and auto quality settings.
Serves both the image generation and image edit endpoints.