An AI-powered platform for creating, editing, and transforming product images into professional marketing materials. Built with FastAPI, React, and powered by cutting-edge AI models including FLUX.1-Kontext for advanced image editing.

• AI Image Generation: Create product images from text descriptions
• Advanced Image Editing: Professional editing with FLUX.1-Kontext
• Color Variations: Generate multiple color variants of products
• Ad Shot Creation: Transform products into surreal advertisement scenes
• Video Animation: Create animated product videos with Kling 2.5

• FLUX.1-Kontext: Advanced image-to-image editing and generation
• FAL Kling 2.5: High-quality video animation generation
• Hunyuan 3D 2.0: 2D to 3D model conversion
• Modal Labs: Scalable GPU compute for AI model deployment

├── app.py                    # Main FastAPI application
├── modal_flux_kontext.py     # FLUX.1-Kontext Modal deployment

├── frontend/
│   ├── index.html           # Main HTML entry point
│   ├── App.js              # React application root
│   ├── components/         # Reusable UI components
│   ├── tools/              # Feature-specific panels
│   └── styles/             # CSS styling

FLUX.1-Kontext is a state-of-the-art image editing model that excels at:

• Context-aware editing: Understanding and preserving image context
• Professional quality: Commercial-grade image transformations
• Flexible prompting: Natural language descriptions for edits
• Style consistency: Maintaining visual coherence across edits

@app.cls(
    image=image,
    gpu="A100-40GB",
    volumes={"/models": volume},
    timeout=900,
    secrets=[modal.Secret.from_name("huggingface")]
)
class FluxKontext:
    def load_model(self):
        self.pipe = FluxKontextPipeline.from_pretrained(
            "black-forest-labs/FLUX.1-Kontext-dev",
            torch_dtype=torch.bfloat16
        )

• Input: Base64 encoded image + text prompt
• Processing: FLUX.1-Kontext applies context-aware transformations
• Output: High-quality edited image maintaining original context

• Color Variations: "Change the product color to red while maintaining lighting"
• Style Transfer: "Make this product look premium and luxurious"
• Background Editing: "Place this product in a modern office setting"
• Material Changes: "Convert this plastic product to metal finish"

@app.post("/api/generate")
async def generate_image(request: GenerateRequest):
    modal_url = os.getenv("MODAL_FLUX_URL")
    
    request_data = {
        "image_base64": request.image_base64,
        "prompt": request.prompt,
        "guidance_scale": request.guidance_scale,
        "num_inference_steps": request.num_inference_steps
    }
    
    async with httpx.AsyncClient(timeout=300.0) as client:
        response = await client.post(f"{modal_url}/generate", json=request_data)

• Python 3.11+
• Node.js 16+
• Modal Labs account
• FAL AI account
• Hugging Face account with FLUX.1-Kontext access

1. Clone the repository

git clone https://github.com/StarAtNyte/VisionCraft.git
cd VisionCraft

2. Install Python dependencies

pip install -r requirements.txt

3. Install Node.js dependencies

npm install

4. Configure environment variables

# Create .env file
FAL_KEY=your_fal_api_key
MODAL_FLUX_URL=your_modal_deployment_url
HF_TOKEN=your_huggingface_token

1. Install Modal CLI

pip install modal
modal setup

2. Deploy FLUX.1-Kontext service

modal deploy modal_flux_kontext.py

3. Get deployment URL

modal app list
# Copy the web app URL to your .env file

1. Get API key from FAL AI
2. Add to .env file as FAL_KEY

# Start main FastAPI server
uvicorn app:app --reload --port 8000

# Start FAL animation service (optional)
python fal_kling_animate.py

# Start Hunyuan 3D service (optional)
cd wan-huggingface && python app.py

# Using Gunicorn
gunicorn app:app -w 4 -k uvicorn.workers.UvicornWorker --bind 0.0.0.0:8000

• Main App: http://localhost:8000
• Studio Interface: http://localhost:8000/
• Marketing Page: http://localhost:8000/home

• Upload product images (JPG, PNG, WEBP)
• Import from URLs
• Drag & drop support

// Example API call
const response = await fetch('/api/generate', {
    method: 'POST',
    headers: { 'Content-Type': 'application/json' },
    body: JSON.stringify({
        image_base64: imageData,
        prompt: "Change the product color to blue while maintaining the original lighting and shadows",
        guidance_scale: 3.5,
        num_inference_steps: 28
    })
});

• Transform products into surreal advertisement scenes
• Context-aware background replacement
• Professional lighting adjustments

• Generate product videos with FAL Kling 2.5

• Multiple animation styles available

• Customizable duration and quality

• Ready for AR/VR applications