Changelog

Reframe is introduced, a new endpoint that enables you to extend and reshape images using generative fill technology. This versatile endpoint offers two distinct modes: directional extension to reveal more content, and aspect ratio adjustment to fit specific canvas dimensions.

Click to visit Developer Documentation

Added a new /account page for easier account management and deletion.

Model Swap is introduced, a new endpoint that enables you to change the identity of fashion models in existing images while preserving clothing and outfit details exactly as they appear. Transform skin tone, facial features, and hair while maintaining the garments, pose, and styling perfectly intact.

Click to visit Developer Documentation

Model Variation is introduced, a new endpoint that enables you to create new variations of existing fashion model images. Transform your images with subtle adjustments or strong modifications while maintaining the core composition.

Click to visit Developer Documentation

Model Create is introduced, a new endpoint that enables you to generate realistic fashion models with simple, intuitive prompts or reference images.

Click to visit Developer Documentation

After creating a model, fine-tune their shots with new extend options:

• Extend Up – Adds more space above the image. Perfect for turning headless shots into full portraits.

• Extend Down – Adds more space below the image. Great for transforming three-quarter shots into full-body views.

• Extend Both – Expands both top and bottom at once, creating a balanced, zoomed-out effect.

A standard set of aspect-ratio badges is introduced in the Gallery and modals, enabling quick reference to the aspect ratio of any image.

Image preview now features a unified blur that preserves aspect ratio across the Application, ensuring a cohesive look and feel.

A number of UX improvements are introduced across the App.

• Drag-and-drop is streamlined: images can be dropped directly onto the "reference image" box in the model-creation flow.

• A responsive grid system is adopted for modals and banners to ensure optimal display on all devices.

• Keyboard navigation is enabled: ←/→ can be used to browse images in Gallery and Model History.

• Scroll position is preserved after image or model deletion, preventing automatic return to the top.

• Empty prompts are supported during model creation; default values are applied to ensure uninterrupted generation.

Model creation now uses our in-house workflow for optimized results with minimal prompting.

• Generate models from a reference image:

  • Pose: Copy the exact pose from an image

  • Silhouette: Preserve body shape and silhouette from an image

• Model generations are automatically saved to /history

• View your previous generations as you continue creating

• The Model Swap feature now runs on FASHN’s in-house solution for faster, more accurate results.

• Effortlessly swap models in any image while keeping clothes, background, and accessories intact.

• Generation time reduced to 12 seconds!

• Try it out on the app: app.fashn.ai/models/change-appearance →

We’ve deployed additional optimizations that reduce end-to-end inference time by approximately 2 seconds. The average runtime now ranges from 10–15 seconds (down from 12–17 seconds), depending on input resolution.

Importantly, this performance boost comes with no loss in output quality.

Resolved a bug where some users were unable to see their generated try-on images in the gallery.

Added aspect ratio selection to the Model Studio, allowing users to choose between formats like 1:1, 4:3, and more when creating a model.

Added GET https://api.fashn.ai/v1/credits to retrieve total, API subscription, and on-demand credits. See docs → https://docs.fashn.ai/fashn-api/endpoints#get-credits-balance