.png)
3 weeks ago
1
🎬 An open-source AI-powered social media platform built with Google's Veo 3.1 video generation and Imagen 4.0 image generation
This project uses Google Cloud Platform's paid APIs. Running this application will incur costs:
- Veo 3.1 Video Generation: ~$0.10-0.30 per video
- Imagen 4.0 Image Generation: ~$0.02-0.05 per image
- Firebase Storage: Pay per GB stored/transferred
- Cloud Firestore: Pay per read/write operation
Estimated costs: $10-50+ per month depending on usage.
💡 Free Development Mode: The backend supports a mock mode (ENABLE_MOCKS=true) for testing without API calls.
The Magic: Your AI Avatar
Upload 3 profile views (front, left profile, right profile) and the app creates a personalized AI model of YOU. Then:
- ✨ Generate videos with yourself in them - "Me skateboarding through cyberpunk Tokyo"
- ✨ Create images featuring you - "Me as a superhero in space"
- ✨ Your likeness is stored in your profile - One-time setup, use everywhere
- ✨ Privacy-first - Your profile images are your data, always under your control
The app automatically passes your profile views to Veo/Imagen when you enable "Include Me" in the composer.
- 🎥 AI Video Generation using Google's Veo 3.1 (up to 8 seconds)
- 🖼️ AI Image Generation using Imagen 4.0
- � Reference Images - Upload up to 3 additional reference images per post (separate from profile)
- Use for style references, objects, or scenes you want in the generation
- Example: Upload a photo of your dog, then generate "my dog surfing in Hawaii"
- 📱 Multiple Aspect Ratios (1:1 square, 9:16 vertical, 16:9 horizontal)
- 🔒 Private/Public Content control
- 📊 Smart Feed Algorithm:
- Interest Feed (60%): Content based on your interactions and preferences
- Explore Feed (25%): Trending and diverse content discovery
- Trending Feed (15%): Popular content across the platform
- ⚡ Real-time Updates with Firebase
- 🎨 Beautiful UI with smooth animations and transitions
- 🔐 Authentication (mock local auth + Firebase Auth ready)
- 📈 Usage Limits to control costs (configurable free tier)
Profile Images vs Reference Images:
- Profile Images (3 views): Your AI avatar, stored in your profile, used when "Include Me" is enabled
- Reference Images (optional, up to 3 per post): Style/object/scene references for individual generations
- Flutter SDK 3.x or higher
- Python 3.11 or higher
- Google Cloud Platform account (for production)
- Firebase project (free tier available)
Backend will be available at http://localhost:8000
For production deployment with real AI generation, see:
- SETUP.md - Complete step-by-step setup
- ARCHITECTURE.md - System design and components
- COSTS.md - Detailed pricing breakdown
- API_REFERENCE.md - API documentation
The backend supports mock mode for development without incurring costs:
Mock mode will:
- ✅ Generate placeholder videos/images instantly
- ✅ Simulate feed algorithms
- ✅ Test all features without API calls
- ✅ Perfect for UI development and testing
- Flutter/Dart - Cross-platform mobile framework
- Riverpod - State management
- Firebase SDK - Auth, Storage, Firestore
- Python 3.11+ - Programming language
- FastAPI - Modern async web framework
- Vertex AI SDK - Google's AI models
- Firebase Admin SDK - Backend services
- Google Cloud Platform:
- Vertex AI (Veo 3.1, Imagen 4.0, Gemini)
- Cloud Firestore (NoSQL database)
- Firebase Storage (Media storage)
- Optional: Cloud Run for deployment
app-demo-short.mp4 app-demo-full.mp4
📱 Smart Feed |
🎥 Post Detail |
👤 Profile |
✏️ Content Composer |
🎨 Generation Options |
⚙️ Settings |
⚙️ Advanced Settings |
🎯 Onboarding |
More Screenshots
 Feed Alternative View |
 Feed Explore Mode |
 Content Browse |
 Post Details |
 Image View |
 Content Grid |
 Generation |
 Results |
 App Interface |
 User Interface |
 Content Creation |
 Video Generation |
 AI Features |
 Profile Setup |
 Content Feed |
 Generation Options |
 Media Gallery |
 Settings Menu |
 User Experience |
 App Navigation |
 Content Discovery |
 Social Features |
 Generation Process |
 Result Viewing |
 Advanced Features |
 Settings Screen |
 Configuration |
 Preferences |
 Advanced Options |
 User Settings |
 App Configuration |
- 🎭 Profile Image Capture: 3-view setup for personalized generation
- 🎬 AI Video Generation: Veo 3.1 creating 8-second videos
- 🖼️ AI Image Generation: Imagen 4.0 with multiple aspect ratios
- 📸 Reference Images: Upload style/object references
- 📊 Smart Feed: Interest-based, Explore, and Trending content
- ⚡ Real-time UI: Smooth animations and transitions
We welcome contributions! Please see CONTRIBUTING.md for guidelines.
- Fork the repository
- Create a feature branch (git checkout -b feature/amazing-feature)
- Make your changes
- Test thoroughly (mock mode + real API if possible)
- Commit your changes (git commit -m 'Add amazing feature')
- Push to your branch (git push origin feature/amazing-feature)
- Open a Pull Request
This project is licensed under the MIT License - see the LICENSE file for details.
- Google Cloud for providing Vertex AI APIs (Veo 3.1, Imagen 4.0, Gemini)
- Flutter Team for the amazing framework
- FastAPI for the excellent Python framework
- Firebase for backend infrastructure
This project is designed for educational and experimental purposes. When deploying:
- Content Moderation: Implement content filtering for user-generated prompts
- Rate Limiting: Control API usage to manage costs
- Terms of Service: Define acceptable use policies
- Privacy: Handle user data responsibly
- Cost Management: Set budget alerts in GCP
- Add web support
- Implement content moderation
- Add video editing features
- Multi-language support
- Advanced feed personalization
- Social features (comments, likes, shares)
- Export/download capabilities
Built with ❤️ for the AI community
⚠️ Reminder: This is experimental software. Use responsibly and monitor your cloud costs!
