Changelog

• Added support for exporting images from the gallery in multiple formats: WebP, PNG, and JPEG.

• Added controls to adjust WebP and JPEG quality for optimized web assets.

Model Swap Improvements

• Model Swap now delivers hyper-realistic results with stronger prompt adherence and seamless blending between the new identity and original clothing.

• Background changes are handled directly in the prompt. If you request a new background, it will be applied; if not, the background remains unchanged.

Backward Compatibility Note

• This upgrade is not yet fully compatible with Consistent Models. If you select a consistent model, the system will fall back to the previous technique so your trained models remain usable.

• Backgrounds will not change by default in this mode. For background edits, we recommend using the Edit tool instead of the old background toggle.

Check it out now: https://app.fashn.ai/models/change-appearance

The features previously found on the Background page are now part of the Edit page, creating a more powerful multi-stage image editor.

You can now edit using prompts, brushes, background removal, and aspect ratio adjustments, while reviewing your results in the history sidebar.

Additional editing tools will be added soon, making the Edit page your complete workspace for preparing images for publication.

Product-to-Model is now accessible through the API. This makes it possible to generate model photos directly from product images in your own applications, with support for the same options available in the app (including model image input, aspect ratio control, and prompt instructions).

Check the docs for implementation details 👉 Product-to-Model API Reference

Previously, the output image automatically matched the aspect ratio of the input product photo. This remains the default, but you can now override it and generate outputs in any aspect ratio you choose, regardless of the original image shape.

The following issues are now fixed:

1. Users couldn't see videos they created from "Product-to-Model" and "Edit" generations.

Users should now be able to see the video they've created retroactively in the Gallery.

2. Users couldn't delete items from Gallery.

• Product-to-Model now accepts a model image in addition to the product image, enabling virtual try-on previews.

• Choose to upload your own image, generate one instantly, or select from your saved models gallery.

• Success rates may vary compared to standard product-to-model outputs, though results can be excellent.

Both prompt-based and brush edits have been upgraded for smarter, more accurate image editing. You can now expect the tool to interpret prompts more effectively, handle brush strokes with greater precision, and produce more realistic results, with less quality degradation.

Try it now in the FASHN AI app 👉 Link

Product-to-Model turns wearable product images into professional model photography. Upload any garment, shoe, or accessory image and generate photos of models wearing the exact item while preserving product details.

The feature analyzes your product image and creates natural model shots with optional customization for person characteristics and settings.

Try it now in the FASHN AI app 👉 Link

Coming soon as an API endpoint

Added a new brush edit option to the image edit page, allowing more precise, prompt-based edits using brush selection.