StyleList: An Open-Source AI Virtual Try-On Tool 70

StyleList is an open-source AI try-on tool that uses computer vision and natural language processing technologies to provide users with personalized clothing recommendations. After users upload a photo, they can receive outfit pairing suggestions for tops and bottoms, and the tool also integrates a product search function. Users simply need to upload a personal photo and a clothing image to generate a preview image of the outfit change, simplifying the process of selecting and styling clothing.

Core Features:

• AI Fashion Stylist: Conversational interface powered by the Llama-4-Maverick-17B model

• Auto-Style Mode: Automatically generates mood boards based on style suggestions, with gender-aware product search

• Image Analysis: Upload photos to receive personalized style recommendations

• Product Search: Amazon product catalog integration via RapidAPI

• Mood Board Creation: Visual collections of selected clothing items, including virtual try-on functionality.

• Multimodal Chat: Supports both text and image input in conversations.

Technical Capabilities Include:

• Streaming AI Responses: Real-time conversation with markdown formatting

• Image Validation: AI-driven photo quality assessment

• Persistent State: Cross-session data storage using Zustand

• Responsive Design: Optimized interfaces for mobile and desktop

Tech Stack:

Frontend:

• Next.js 15 (with App Router) and TypeScript

• React 18 (with server and client components)

• TailwindCSS (for styling and responsive design)

• Zustand (for state management with persistence)

• ReactMarkdown (for formatting AI responses)

• Lucide Icons (for UI icons)

Backend & API:

• Vercel AI SDK (for streaming AI interactions)

• Amazon Product API (product data via RapidAPI)

• Custom API Routes (for chat, image validation, and mood board generation)

• Server Actions (for form handling and data processing)

AI Integration:

• Llama-4-Maverick-17B-128E-Instruct-FP8 model (provides fashion expertise)

• Vision capabilities for image analysis and style recommendations

• Tool calling for product search integration

• Custom system prompts optimized for fashion styling

Installation & Configuration:

• Prerequisites: Node.js 18.0 or higher, npm or yarn package manager, a RapidAPI account for Amazon product access.

• Installation Steps:

1. Clone the repository: git clone <repository-url>, then navigate to the project directory: cd open-ai-stylist.

2. Install dependencies: npm install.

3. Configure environment variables (create a .env.local file).

4. Start the development server: npm run dev.

5. Open http://localhost:3000 in your browser.

How to Use:

1. Initial Setup: Upload an approved model photo. The photo will be validated by AI to ensure it is suitable for styling, and approved photos will be stored for styling conversations.

2. Styling Conversation: Start a new chat session, upload a photo or describe your styling needs, receive personalized recommendations for tops and bottoms, browse suggested products with real-time Amazon pricing, and select items to create a mood board.

3. Mood Board Panel: Select products in the conversation, create themed collections with AI-generated titles, view products in a responsive grid layout, and access direct purchase links for selected items.

4. Auto-Style Mode: Click the lightning button in the chat input area to enable auto-styling. Converse normally with the AI stylist. When the AI provides styling suggestions, auto-styling will analyze the suggestions to detect gender context, extract key clothing items mentioned, perform gender-aware product searches, generate virtual try-ons using your model photo, and automatically create a complete mood board. You will be notified upon completion.

Use Cases

• Daily Outfit Planning: Helps you try different clothing combinations from home, making it convenient for everyday wear.

• Apparel E-commerce Assistance: Apparel e-commerce sellers can leverage this tool to provide virtual try-on services for buyers, increasing conversion rates.

Open-Source Address:

https://github.com/parsakhaz/open-ai-stylist