How to Integrate GPT Image 2.6 with ModelFold
ModelFold published a guide for calling six GPT Image 2.6 variants through one API key and one image-generation endpoint, covering text-to-image and image-to-image tasks across Economy, Flare, and Sunburst tiers. The guide documents model IDs, request parameters, 1K–4K resolution options, reference-image inputs, and GIF output, while a web Playground lets users inspect forms and examples before coding. These availability and capability details are vendor claims based on ModelFold model pages and API documentation.
The documented endpoint, model IDs, and parameters give developers a concrete path to integrate GPT Image 2.6 for text-to-image, image editing, multiple resolutions, and GIF generation through ModelFold.
GPT Image 2.6 可以在 ModelFold 里直接调用。一个 API Key，同一个接口，用六个模型 ID 区分档位和任务。网页上的 Playground 能先看表单和示例，再按同样的参数写进代码。
本文由 ModelFold 团队编写，只写接入方法。能力描述来自 2026 年 11 月 7 日的模型页和接口文档。页面以后若有改动，以页面为准。
六个模型 ID
| 模型 ID | 用途 |
|---|---|
| gpt-image-2-6/economy-text-to-image | Economy 文生图 |
| gpt-image-2-6/economy-image-to-image | Economy 图生图 |
| gpt-image-2-6/flare-text-to-image | Flare 文生图 |
| gpt-image-2-6/flare-image-to-image | Flare 图生图 |
| gpt-image-2-6/sunburst-text-to-image | Sunburst 文生图 |
| gpt-image-2-6/sunburst-image-to-image | Sunburst 图生图 |