Skip to content

symesj/avatar-3d

 
 

Repository files navigation

Avatar 3D

Generate interactive 3D avatars from a single photo using AI. Choose between cursor-tracking rotation or full 3D model export.

Features

  • Two Generation Modes:
    • Cursor Tracking - Generate a grid of head rotation frames for mouse-follow effect
    • 3D Model - Generate an actual GLB 3D model you can use in Blender, Unity, etc.
  • Pixar-Style Preprocessing - Photos are transformed into stylized 3D characters
  • Custom Style Prompts - Add accessories, change styles (cyberpunk, cartoon villain, etc.)
  • Export Options - Download GLB files, HTML embeds, individual frames, or React components

Inspiration

Inspired by Wes Bos's Eye Ballz project, which creates interactive eye-tracking avatars. This project extends the concept with AI-powered stylization and true 3D model generation.

How It Works

  1. Upload a photo with a face
  2. The app preprocesses your photo using google/nano-banana-pro to create a Pixar-style 3D character
  3. Based on your selected mode:
  4. Preview and export your creation

Getting Started

Prerequisites

Installation

# Clone the repo
git clone https://github.com/0xGF/avatar-3d.git
cd avatar-3d

# Install dependencies
pnpm install

# Copy environment variables
cp .env.example .env.local

# Add your Replicate API token to .env.local
# REPLICATE_API_TOKEN=r8_xxxxx

# Run the dev server
pnpm dev

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

Configuration

Cursor Tracking Mode

  • Horizontal/Vertical Steps: Number of images in each direction (5x5 = 25 frames)
  • Higher step counts = smoother rotation but more API calls

3D Model Mode

  • Texture Size: Resolution of the model texture (512-2048px)
  • Mesh Quality: Detail level of the 3D mesh (50-100%)

Cost Estimates

  • Preprocessing: ~$0.01 per image
  • Cursor Tracking: ~$0.01 per frame (5x5 grid ≈ $0.25)
  • 3D Model: ~$0.05-0.10 per generation

Tech Stack

Star History

Star History Chart

License

MIT

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages

  • TypeScript 95.1%
  • CSS 4.3%
  • JavaScript 0.6%