Turn any GitHub repository into a viral promo video in just one-click.
RepoToViralVideo analyzes any GitHub repository and generates a short, high-energy promo video designed to go viral on X/LinkedIn with AI voiceover, kinetic typography, animated stats, and background music.
⚡ Powered by Gemini 3.1 Pro (
gemini-3.1-pro-preview) — Google's latest model for deep repo analysis, feature extraction, and engaging script generation.
No video editing skills needed. Just point it at a repo and get a ready-to-share video.
- 🔥 Kinetic Typography: Words slam in from all directions with spring physics
- 📊 Animated Counters: Star counts and fork counts animate from 0 to their final number
- 🎤 AI Voiceover: Gemini TTS generates natural, conversational narration for each scene
- 🎵 Background Music: 4 bundled royalty-free tracks (chill, upbeat, tech, hype) or use your own
- ⚡ Fast Cuts: Slide transitions between scenes, nothing stays static
- 🧠 Smart Analysis: Gemini 3.1 Pro reads the repo page directly and extracts impressive stats, features, and selling points
- 🎯 Adaptive Scenes: Automatically selects different video styles based on repo maturity (4–6 scenes)
- 🖥️ Web UI: Modern frontend to generate videos right from your browser
Here's a viral video generated for awesome-llm-apps (95K+ ⭐):
- Python 3.10+
- Node.js 18+
- FFmpeg:
brew install ffmpegon macOS orapt install ffmpegon Ubuntu - Gemini API key: Get your API from Vertex AI Studio or Google AI Studio.
# 1. Clone the repo
git clone https://github.com/Shubhamsaboo/repotovideo.git
cd repotovideo
# 2. Install all dependencies (Python + Node)
npm run install:all
# 3. Set your Gemini API key
export GEMINI_API_KEY="your-key-here"This installs Python dependencies, the video renderer, and the web frontend in one command.
python generate.py https://github.com/user/cool-repoThis will analyze the repo, generate a voiceover, and render a 1080p video to apps/video/out/viral-<repo-name>.mp4.
You can customize the output with flags:
python generate.py https://github.com/user/repo --music hype --voice PuckMusic moods: tech (default), hype, chill, upbeat
Voices: Puck (default, energetic), Kore (confident), Aoede (warm), Charon (deep)
# Terminal 1 — Start the API server (from project root)
cd apps/api
python3 server.py
# Terminal 2 — Start the web frontend (from project root)
npm run dev:webOpen http://localhost:3000, enter your Gemini API key and a GitHub URL, pick your music and voice, and hit Generate. The API runs on port 8000 and the frontend on port 3000.
The pipeline runs in 3 steps:
- AI Analysis & Script: Gemini 3.1 Pro reads the repo page directly (via URL context), extracts stars, forks, features, and tech stack, then writes a punchy Fireship-style voiceover script with scene selection
- Generate Voiceover: Gemini 2.5 Pro TTS creates natural narration with per-scene tone guidance ("say this with building excitement")
- Render Video: Remotion renders dynamic React components into a 1080p MP4 with spring animations, transitions, and background music
The AI adapts the video based on repo maturity:
| Repo Size | Stars | Scenes | Hook Style |
|---|---|---|---|
| 🚀 Viral | 10K+ | 6 | Animated star counter |
| 📈 Growing | 100–10K | 5 | Momentum ("X K+ and climbing") |
| 💡 New | <100 | 4 | Problem-focused |
All voices use Gemini 2.5 Pro TTS with conversational tone prompting:
| Voice | Style |
|---|---|
| Puck ⭐ | Playful, energetic: great for hype videos |
| Kore | Warm, confident |
| Aoede | Smooth, warm |
| Charon | Deep, authoritative |
repotovideo/
├── generate.py # CLI entry point
├── apps/
│ ├── api/ # Python FastAPI backend
│ │ ├── server.py # API server (port 8000)
│ │ ├── requirements.txt
│ │ └── pipeline/
│ │ ├── viral_analyzer.py # Gemini 3.1 Pro repo analysis + URL context
│ │ ├── viral_tts.py # Gemini 2.5 Pro TTS voiceover
│ │ └── viral_renderer.py # Dynamic Remotion composition generator
│ ├── video/ # Remotion video renderer
│ │ ├── src/
│ │ │ ├── TutorialVideo.tsx # Main composition (auto-generated)
│ │ │ ├── Root.tsx # Remotion root (auto-generated)
│ │ │ ├── types.ts # RepoData interface
│ │ │ └── scenes/ # Scene components
│ │ │ ├── HookScene.tsx # Star counter / momentum / problem hook
│ │ │ ├── WhatScene.tsx # Kinetic title animation
│ │ │ ├── FeaturesScene.tsx
│ │ │ ├── TechScene.tsx
│ │ │ ├── StatsScene.tsx
│ │ │ └── CTAScene.tsx
│ │ └── public/music/ # Bundled royalty-free tracks
│ └── web/ # Next.js frontend
│ ├── src/app/page.tsx # Landing page + generate UI
│ └── public/assets/ # Visual assets + demo video
- AI: Gemini 3.1 Pro (analysis) + Gemini 2.5 Pro TTS (voiceover)
- Video: Remotion (React → MP4)
- Backend: FastAPI (Python)
- Frontend: Next.js + Framer Motion
- Audio: FFmpeg (WAV → MP3 conversion)
MIT - fork it, customize it, make it yours.