Skip to content

Clothing AI Try-On

AI-powered virtual try-on for clothing using digital twin technology.

What It Does

Shoppers can see how clothing looks on them (or a model) using generative AI. The system creates a digital twin of the person and applies selected garments virtually.

Key Features

  • Digital twin — AI-generated representation of the shopper
  • Virtual fitting — see how clothes look on the digital twin
  • Size fitting — visualization of size recommendations (new, not fully public yet)
  • Embeddable — clients can embed the try-on in an iframe or modal window

How to Use

Method Description
Frontend app tryon.wearfits.com — example implementation
Iframe / modal Embed tryon.wearfits.com in your website
API Direct API calls for custom implementations — see API Reference

API Endpoints

Endpoint Description
POST /digital-twin Generate a digital twin image
POST /virtual-fitting Apply clothing to digital twin

Domain

How It Works

sequenceDiagram
    participant User
    participant App as tryon.wearfits.com
    participant API as api.wearfits.com
    participant Modal as Modal.com (GPU)

    User->>App: Capture selfie + enter measurements
    App->>API: POST /api/v1/digital-twin
    API->>Modal: Generate body mesh (SAM3D/MHR)
    Modal-->>API: Digital Twin ID
    API-->>App: Job completed (digitalTwinId)
    User->>App: Select garments
    App->>API: POST /api/v1/virtual-fitting
    API->>Modal: AI rendering
    Modal-->>API: Result image
    API-->>App: Result URL
    App->>User: Display fitting result

Digital Twin Modes

Mode Input Description
Size Mode Face photo + height + clothing size (XS–3XL) Default mode — measurements-based body model
Photo Mode Face photo + full-body photo Body mesh extracted from photo via SAM3D

Authentication

  • Frontend app (tryon.wearfits.com): Protected by Cloudflare Turnstile (bot protection), no API key needed for end users
  • Direct API calls: Require X-API-Key header with key from dash.wearfits.com

Rate Limits

Context Limit
Frontend app (per user) 10 fittings/hour, 20/day
API (per key, job submission) 50/min
API (job status polling) 500/min

Further Reading