FASHN, in collaboration with fal, has introduced an advanced virtual try-on model, marking a significant milestone as the first commercially available solution of its kind. This sophisticated pipeline comprises seven models, with a central 1.2 billion-parameter flow-matching model trained on a dataset of 4 million meticulously curated images, ensuring precise garment detail preservation at a resolution of 576x864 pixels. Designed for versatility, the model supports a wide range of clothing categories but excludes intimate apparel and swimwear to minimize NSFW biases. Its capabilities include accommodating various starting outfits and poses, with optimal performance when images are well-aligned, and providing fast processing times of approximately 15 seconds. The model excels in delivering pixel-perfect accuracy, particularly with graphic prints, and offers extensive customization options such as background restoration and sampling controls. Users can engage with the model through a web app interface that integrates AI-driven innovations like FLUX, enhancing the virtual try-on experience, and can find further information through FASHN's FAQ, Roadmap, and Usage Guides.