Out-of-the-box, High-Performance Full-Stack Open-Source AI Video & Image Generation Platform
🚀 Live Demo • ⚡ Quick Start • ✨ Features • ⚙️ Integrations • 📚 Docs
🚀 Battle-Tested in Production: VidGen powers the commercial platform vidgenerator.ai.
VidGen is an open-source, production-ready AI video and image generation platform designed for creators, indie hackers, and entrepreneurs.
Whether you want to launch a full-featured Midjourney / Luma alternative in 5 minutes or build a custom AI monetization business, VidGen provides a complete end-to-end solution: from a modern SSR Nuxt 3 frontend portal, bilingual admin dashboard, and Celery async generation engine, to credit economics, payment gateways, prompt community, virtual user generators, and content moderation.
No need to install Python/Node.js, configure databases, or even apply for paid API keys (built-in Mock generation mode and demo data for zero-cost local testing). Run the entire stack with a single command!
# 1. Clone repository
git clone https://github.com/ucmao/vidgen.git
cd vidgen
# 2. Launch all 6 services with Docker Compose (Web, Admin, FastAPI, Celery, Postgres, Redis)
docker compose up -dOnce started, open your browser:
| Service | URL | Default Credentials / Note |
|---|---|---|
| User Web Portal | http://localhost:3000 |
Preloaded with demo data & Mock AI engine for instant testing |
| Admin Panel | http://localhost:3001 |
Username: admin (check terminal logs for the generated password, or set INITIAL_ADMIN_PASSWORD) |
| Interactive API Docs | http://localhost:8000/docs |
Swagger UI documentation & testing |
💡 Quick Tips:
- View real-time startup logs:
docker compose logs -f backend.- On initial container launch, database migrations and demo seed data are automatically applied.
- Text-to-Image (Text2Img): Generate high-resolution visuals via FLUX.1, SDXL, Midjourney, and more.
- Image-to-Image (Img2Img): Image-guided synthesis with prompt tuning and strength control.
- Text-to-Video (Text2Video): Turn prompts into dynamic AI video clips (HunyuanVideo, Luma, Pika, Runway).
- Image-to-Video (Img2Video): Animate still photos into high-definition videos with motion control.
- Global Payments: Native integration with PayPal and Stripe for subscriptions and credit top-ups.
- Granular Credit Pricing: Set custom credit deductions per model, resolution, and step.
- Growth & Marketing: Promo codes, discount campaigns, and tiered daily check-in streak rewards.
- 1-Click Bilingual Toggle: Switch seamlessly between English and Simplified Chinese (
English/中文). - Sockpuppet / Virtual User Generator: Generate realistic synthetic user accounts & avatars to jumpstart community engagement.
- Content Moderation: Automated NSFW filtering, sensitive word blocklist, report handling, and user ban tools.
- Dynamic Pricing Manager: Modify model status, pricing, and multipliers on the fly without redeploying.
- Explore Gallery: Responsive masonry layout with 1-click prompt copying and parameter inspection.
- Creator Profiles: Custom handles (
@username), bios, avatars, and personal work showcases. - Social Engagement: Likes, favorites, comments, and creator follows.
- SEO Ready: Auto-generated sitemaps, category aggregation (
/category/...), and effect pages (/effects/...).
- Celery Task Queues: Heavy AI generation jobs run completely asynchronously without blocking API threads.
- Real-Time WebSocket: Instant job completion pushes directly to the frontend.
Ready to move from local testing to real AI generation? Simply fill in your API keys in backend/.env (or Docker environment variables):
# 1. Real AI Providers (Setting a key automatically disables Mock mode)
REPLICATE_API_KEY="r8_your_replicate_api_key_here"
# GEMINI_API_KEY="your_gemini_api_key_here"
# 2. Payment Gateways (PayPal / Stripe)
PAYPAL_CLIENT_ID="your_paypal_client_id"
PAYPAL_CLIENT_SECRET="your_paypal_client_secret"
# STRIPE_SECRET_KEY="sk_live_..."
# 3. SMTP Email (For verification codes)
SMTP_HOST="smtp.example.com"
SMTP_PORT="465"
SMTP_USER="noreply@example.com"
SMTP_PASSWORD="your_smtp_password"
# 4. Google OAuth Login (Optional)
GOOGLE_CLIENT_ID="your_google_client_id"
GOOGLE_CLIENT_SECRET="your_google_client_secret"| Component | Technologies |
|---|---|
| Frontend Web | Nuxt 3.21.11 (Vue 3, SSR/ISR) + Tailwind CSS + Pinia + Lucide Icons |
| Admin Panel | Nuxt 3.21.11 (Vue 3) + Tailwind CSS + Custom Bilingual i18n (EN/ZH) |
| Backend API | FastAPI 0.141.1 (Python 3.11+) + SQLAlchemy 2.0 + Pydantic 2 |
| Task Queue & Cache | Celery 5.4+ + Redis 7 (Broker & Cache) + Flower (Queue Monitor) |
| Database | PostgreSQL 15+ + Alembic Migrations |
| Storage & CDN | Cloudflare R2 / AWS S3 / Aliyun OSS (S3-Compatible Object Storage) |
| Deployment | Docker Compose / systemd Automation Scripts / Nginx Reverse Proxy |
For manual development setup, architectural deep-dives, or production deployment guides, check out the documentation library:
- 🚀 Getting Started:
- 🏗️ Architecture & Systems:
- 🔌 Integrations:
- 🚢 Production Deployment:
To deploy VidGen in production, use docker-compose.prod.yml with a private .env.production configuration:
# Generate strong secrets and credentials in .env.production, then run:
docker compose --env-file .env.production -f docker-compose.prod.yml up -dFor complete production security baselines and TLS setup, see Production Deployment Docs.
This project is licensed under the MIT License — free for both personal and commercial use.
- Submit an Issue to report bugs or request features.
- Pull Requests are always welcome!
- If VidGen helps you, please consider giving us a ⭐️ Star on GitHub!