An open-source, AI-native video editor. Edit with natural language, search your footage semantically, and export in 4K — all in the browser.
Automate video editing workflows using conversational AI agents, smart cuts, and translation.
| Name | Description | Status |
|---|---|---|
| Conversational AI Director | Edit videos via chat commands. | ✓ |
| Edit for Clarity | Auto-remove filler words and silences. | ✓ |
| Viral Clip Creation | Package engaging moments into portrait clips. | ✓ |
| Highlight Reels | Auto-generate supercuts of footage. | ✓ |
| Auto-Chapters | Automatically segment video with timestamps. | ✓ |
| AI Translation & Dubbing | Translate transcripts and voice tracks. | ✓ |
| Automatic Multicam | Switch camera angles based on speaker activity. | In-Progress |
Search and index your video library semantically with vision-language models and pgvector RAG.
| Name | Description | Status |
|---|---|---|
| Visual & Transcript Indexing | Extract visual descriptions and transcripts. | ✓ |
| pgvector Similarity Search | Run semantic queries across the library. | ✓ |
| Scene-Level Deep Linking | Jump directly to matched event frames. | ✓ |
A professional browser-based multi-track timeline editor powered by GPU shaders and keyframe animations.
| Name | Description | Status |
|---|---|---|
| Multi-Track Editing | Support video, audio, overlay, and caption tracks. | ✓ |
| Dynamic Snap & Ripple Trim | Drag-and-drop with snapping and ripple trim. | ✓ |
| 50+ Custom GPU Shaders | Real-time transitions and GLSL visual filters. | ✓ |
| Keyframe-Accurate Animation | Animate position, scale, opacity, and rotation. | ✓ |
| Interactive Word-by-Word Captions | Synced captions with word-level styling. | ✓ |
Advanced visual effects and generative AI tools for gaze correction, background removal, and media generation.
| Name | Description | Status |
|---|---|---|
| AI Eye Contact Correction | Realign speaker's gaze to the camera. | ✓ |
| Smart Background Removal | Virtual green screen background replacement. | ✓ |
| Active Speaker Tracking | Auto-crop and center active speaker. | ✓ |
| Generative Media | Text-to-Image/Video via serverless GPU workflows. | ✓ |
Deep-learning audio processors to enhance voice quality, remove silences, and detect filler words.
| Name | Description | Status |
|---|---|---|
| Studio Sound | Enhance voice and eliminate background noise. | ✓ |
| Smart Silence Compressor | Compress long pauses and awkward gaps. | ✓ |
| Filler Word Detection | Clip out verbal clutter directly from timeline. | ✓ |
Hardware-accelerated rendering and high-fidelity video exports directly in the browser sandbox.
| Name | Description | Status |
|---|---|---|
| WebCodecs Acceleration | Hardware-accelerated browser decoding/encoding. | ✓ |
| PixiJS Compositor | Sub-frame accurate canvas-based rendering engine. | ✓ |
| 4K Exports | High-fidelity exports in the browser sandbox. | ✓ |
apps/
├── app/ # Next.js 15 — frontend, auth
└── docs/ # Documentation
services/
├── director/ # NestJS Fastify — REST API, WebSocket gateway, BullMQ workers
├── modal-processor/ # Video conforming & asset pipeline (Python, Modal GPU)
├── modal-video-renderer/ # Server-side video rendering tasks
└── modal-media-generator/ # Image / video / audio generation (ElevenLabs, Kling)
packages/
├── db/ # Drizzle ORM + PostgreSQL schema (shared)
├── api/ # tRPC routers (type-safe RPC layer)
├── auth/ # Better Auth (magic link + GitHub OAuth)
├── engine-pixi/ # PixiJS compositor & WebCodecs export engine
├── timeline/ # Timeline UI component & track manager
└── core/ # Shared schemas, commands, playback state
Key integrations: PostgreSQL · pgvector · Redis (BullMQ) · Cloudflare R2 · Deepgram · Gemini · ElevenLabs · Modal · TwelveLabs
git clone https://github.com/openvideodev/openvideo.git
cd openvideo
pnpm install
cp apps/app/.env.sample apps/app/.env.local
cp services/director/.env.example services/director/.env
# Apply database migrations
pnpm --filter @openvideo/db db:push
# Start all dev servers
pnpm devApp → http://localhost:3000 · Director API → http://localhost:4000
See LICENSE.