AI-Powered 3D Product & Packaging Design Platform
Transform ideas into production-ready 3D models and custom packaging in minutes.

CRAFTO is an end-to-end AI design platform that bridges the gap between imagination and production. Using cutting-edge generative AI, it transforms simple text descriptions into:

• 3D Product Models – Generate photorealistic 3D assets from text or images
• Custom Packaging – Design professional packaging with AI-generated panel textures
• Interactive Previews – Real-time 3D visualization with lighting and material controls

Whether you're a designer prototyping concepts, a marketer creating product mockups, or an entrepreneur visualizing ideas – CRAFTO makes professional 3D design accessible to everyone.

• Text-to-3D Pipeline: Describe any product and watch it materialize in 3D
• Multi-Angle Synthesis: AI generates multiple views for accurate 3D reconstruction
• Iterative Editing: Refine models with natural language instructions
• Export Ready: Download production-ready GLB files

• Parametric Templates: Box and cylinder packaging with custom dimensions
• Industry Presets: Pre-configured sizes for common products (shoes, bottles, electronics)
• Panel-by-Panel AI Generation: Generate unique textures for each packaging face
• Live Dieline Editor: Visual feedback with UV-mapped texture placement
• 3D Preview: Real-time package visualization with product integration

• Trellis 3D Integration: State-of-the-art image-to-3D reconstruction via fal.ai
• Gemini Vision AI: Multi-modal image generation with style consistency
• Context-Aware Editing: AI maintains design coherence across iterations
• Caching System: Smart asset caching for instant load times
• Redis State Management: Seamless session persistence and recovery

┌─────────────────────────────────────────────────────────────┐
│                     CRAFTO Platform                          │
├─────────────────────────────────────────────────────────────┤
│                                                               │
│  Frontend (Next.js 15 + TypeScript + Three.js)              │
│  ├─ Product Creation Flow                                    │
│  ├─ Packaging Design Studio                                  │
│  ├─ 3D Model Viewer (React Three Fiber)                     │
│  └─ Real-time Chat Interface                                 │
│                                                               │
├─────────────────────────────────────────────────────────────┤
│                                                               │
│  Backend (FastAPI + Python)                                  │
│  ├─ Product Pipeline Service                                 │
│  │   ├─ Gemini Image Generation (Multi-View)               │
│  │   └─ Trellis 3D Model Synthesis                         │
│  │                                                            │
│  ├─ Packaging Generation Service                             │
│  │   ├─ Panel Texture Generation                            │
│  │   ├─ Dieline Processing                                  │
│  │   └─ UV Mapping & Export                                 │
│  │                                                            │
│  └─ State Management (Redis)                                 │
│                                                               │
├─────────────────────────────────────────────────────────────┤
│                                                               │
│  AI Services                                                  │
│  ├─ Google Gemini 3 Image (Multi-angle views)               │
│  ├─ Trellis 3D Reconstruction (fal.ai)                      │
│  └─ Context-aware prompt engineering                         │
│                                                               │
└─────────────────────────────────────────────────────────────┘

Tech Stack:

• Frontend: Next.js 15, React Three Fiber, Radix UI, TailwindCSS
• Backend: FastAPI, Pydantic, Redis
• AI: Google Gemini 3 Image, Trellis (fal.ai)
• 3D: Three.js, GLB/GLTF, trimesh
• Infrastructure: Docker, Docker Compose

• Node.js 18+ and npm/yarn
• Python 3.11+
• Redis (local or cloud instance)
• API Keys:
  • Google Gemini API Key (Get one here)
  • fal.ai API Key (Get one here)

1. Clone the repository

git clone https://github.com/yourusername/CRAFTO.git
cd CRAFTO

2. Backend Setup

cd backend

# Create virtual environment
python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

# Install dependencies
pip install -r requirements.txt

# Configure environment
cp env.example .env
# Edit .env with your API keys:
#   FAL_KEY=your_fal_api_key
#   GEMINI_API_KEY=your_gemini_api_key
#   REDIS_URL=redis://localhost:6379/0

# Start the backend
uvicorn main:app --host 0.0.0.0 --port 8000 --reload

3. Frontend Setup

cd frontend

# Install dependencies
npm install

# Start development server
npm run dev

4. Access the Application

• Frontend: http://localhost:3000
• Backend API: http://localhost:8000
• API Docs: http://localhost:8000/docs

1. Navigate to Home: Enter a product description

   "A modern ceramic coffee mug with a matte finish"
   

2. Generate: Click "Create" and watch the AI work:

   • 🖼️ Generates multi-angle product views
   • 🎨 Reconstructs 3D geometry from images
   • ✨ Produces a downloadable 3D model

3. Refine: Use the chat panel to edit:

   "Make it blue with gold trim"
   "Add a geometric pattern"
   

1. Select Package Type: Choose box or cylinder
2. Set Dimensions: Use presets or custom measurements
3. Generate Textures: Describe each panel:

   "Minimalist design with brand logo and product showcase"
   

4. Preview in 3D: See your packaging in real-time
5. Export: Download UV maps and 3D models

Backend (backend/.env):

# Required
FAL_KEY=your_fal_api_key_here
GEMINI_API_KEY=your_gemini_api_key_here
REDIS_URL=redis://localhost:6379/0

# Optional - Model Selection
GEMINI_PRO_MODEL=gemini-3-pro-image-preview
GEMINI_FLASH_MODEL=gemini-2.5-flash-image
GEMINI_IMAGE_SIZE=1K  # Options: 1K, 2K, 4K

# Optional - Quality Settings
SAVE_ARTIFACTS_LOCALLY=false  # Save all generations locally
DEMO_MOCK_MODE=false          # Enable for presentations (no API calls)

Frontend (frontend/next.config.ts):

// API endpoints auto-configured for localhost development
// Update for production deployment

Run the entire stack with Docker Compose:

cd backend

# Configure .env file first
cp env.example .env
# Edit .env with your API keys

# Start all services
docker-compose up -d

# View logs
docker-compose logs -f

# Stop services
docker-compose down

Services:

• Backend API: http://localhost:8000
• Frontend: Configure separately or add to compose

CRAFTO/
├── backend/                  # FastAPI backend
│   ├── app/
│   │   ├── core/            # Config, Redis, utilities
│   │   ├── endpoints/       # API routes
│   │   │   ├── product/     # Product generation API
│   │   │   ├── packaging/   # Packaging design API
│   │   │   └── trellis/     # 3D generation API
│   │   ├── integrations/    # External AI services
│   │   │   ├── gemini.py    # Google Gemini integration
│   │   │   └── trellis.py   # Trellis 3D integration
│   │   ├── models/          # Pydantic models
│   │   └── services/        # Business logic
│   ├── tests/               # Test suite
│   ├── main.py             # FastAPI app entry
│   └── requirements.txt    # Python dependencies
│
├── frontend/                # Next.js frontend
│   ├── app/                # App router pages
│   │   ├── product/        # Product creation flow
│   │   ├── packaging/      # Packaging designer
│   │   └── page.tsx        # Home page
│   ├── components/         # React components
│   │   ├── ModelViewer.tsx
│   │   ├── package-viewer-3d.tsx
│   │   ├── dieline-editor.tsx
│   │   └── AIChatPanel.tsx
│   ├── lib/               # API clients & utilities
│   │   ├── product-api.ts
│   │   ├── packaging-api.ts
│   │   └── model-cache.ts
│   └── package.json
│
└── docs/                   # Documentation
    ├── Packaging.plan.md
    └── product_backend_plan.md

• POST /product/create – Generate new 3D product
• POST /product/edit – Refine existing product
• POST /product/trellis-only – Convert pre-generated images to 3D
• GET /product – Get current product state
• GET /product/status – Poll generation status

• POST /packaging/initialize – Start packaging session
• POST /packaging/panels/generate – Generate panel texture
• POST /packaging/panels/bulk-generate – Generate multiple panels
• GET /packaging/state – Get current packaging state
• POST /packaging/export – Export packaging files

• GET /health – Health check
• GET / – API welcome message

Full API documentation: http://localhost:8000/docs

cd backend

# Run all tests
pytest

# Run with coverage
pytest --cov=app tests/

# Run specific test file
pytest tests/test_product_pipeline.py -v

• Text-to-3D product generation
• Iterative AI editing
• Box and cylinder packaging
• Panel-by-panel texture generation
• Real-time 3D preview

We welcome contributions! Here's how:

1. Fork the repository
2. Create a feature branch (git checkout -b feature/amazing-feature)
3. Commit changes (git commit -m 'Add amazing feature')
4. Push to branch (git push origin feature/amazing-feature)
5. Open a Pull Request

Please ensure:

• Code follows existing style conventions
• Tests pass (pytest for backend, npm test for frontend)
• Documentation is updated

• Google Gemini – Multi-modal AI image generation
• Trellis (fal.ai) – State-of-the-art image-to-3D reconstruction
• Three.js – 3D rendering engine
• FastAPI – High-performance Python web framework
• Next.js – React framework for production

• Issues: GitHub Issues
• Discussions: GitHub Discussions