> For the complete documentation index, see [llms.txt](https://docs.styleai.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.styleai.io/product-docs/product-docs-en/readme-1/26.07.06-gpt-image-2.md).

# 26.07.06 GPT Image 2

An AI model that is somewhat slow, but delivers overwhelming material texture rendering!

The newly introduced GPT 2.0, in order to express the fine texture of fabrics \
**An average of 120 to 250 seconds**of precise computation time is required.&#x20;

#### 1. GPT 2.0 model features (High-Quality & Slow)

* Fabric texture depiction: It precisely reproduces the texture of difficult fabrics, such as the fine grain of leather and the three-dimensional structure of tweed or knitwear, just like real clothes.
* Detail-focused computation: It renders each element by calculating the natural drape of the garment and even the fine seams one by one.

| Fabric                                                                                                | Nano Banana Pro                                                                         | GPT Image 2                                          |
| ----------------------------------------------------------------------------------------------------- | --------------------------------------------------------------------------------------- | ---------------------------------------------------- |
| <p><img src="/files/bf2414984a61ac9efcf78633142aaaa217c8a99d" alt="" data-size="original"></p><p></p> | <img src="/files/147345b19e64807ede8b3d2b12af7f30d6d0c768" alt="" data-size="original"> | ![](/files/3ab968b6975fca80a1c602c94271ed305b6d37fd) |

#### ⏱ Wait time policy update notice (Max 5 Min)

Because GPT 2.0 is relatively slower at generating than other models, the system policy has been changed as follows to alleviate designers' frustration and provide smoother service.

* 5-minute generation wait time (timeout) applied: The maximum generation wait time has been adjusted to 5 minutes. Jobs that are not completed even after 5 minutes will be automatically canceled (failed).

When you need definite 'one image quality' even if it's a little slow, try selecting GPT 2.0 in the image generation menu now!
