Positioning and release
GPT-Image-2.5 Flare is OpenAI’s speed-oriented image generation and editing model for everyday applications. OpenAI announced it on September 8, 2026. The dated API snapshot is gpt-image-2.5-flare-2026-09-08. (openai.com)
Capabilities and uses
- Accepts text and image inputs and produces images. It can be selected in the Images API or as the image-generation model in the Responses API. (developers.openai.com)
- Supports image generation, reference-based editing, transparent backgrounds, and
auto,low,medium,high,xhigh, andmaxquality settings. (developers.openai.com) - Intended uses include creator and social content, product experiences, visual search, rapid prototyping, and high-volume image generation. At launch, OpenAI reported 50% lower latency than GPT-Image-2; actual results depend on the workload. (openai.com)
Limitations
GPT Image models can struggle with exact text placement, consistent recurring characters or brand elements, and precise composition. Prompts and generated images are subject to content filtering. Flare does not support audio or video; the model documentation also lists fine-tuning, function calling, and structured outputs as unsupported. (developers.openai.com)
Custom image dimensions must be multiples of 16, with neither edge above 3,840 pixels, an aspect ratio no greater than 3:1, and a total of 655,360–8,294,400 pixels. Outputs above 3,686,400 pixels are experimental. (developers.openai.com)
API pricing
Currency: USD; rates per 1 million tokens. Official prices checked on November 16, 2026. (platform.openai.com)
| Processing | Text input | Image input | Image output |
|---|---|---|---|
| Standard | $5.00 | $8.00 | $30.00 |
| Batch API | $2.50 | $4.00 | $15.00 |
For standard processing, cached text input is $1.25 and cached image input is $2.00 per 1 million tokens. For GPT-Image-2.5, cached-input pricing applies only to the image-generation tool in the Responses API, not direct Images API requests. The model outputs images rather than text, so there is no text-output charge. Responses API requests also incur charges for the mainline model’s token usage. Actual image cost varies with token consumption, including the selected size and quality. (developers.openai.com)