AI Meeting Assistant & Real-Time Interview Copilot
NexQ overlay during a live interview — real-time transcription and AI suggestions
🔒 100% Local — your audio and data never leave your machine
🆓 Free & Open Source — no subscriptions, no limits, ever
⚡ 10 STT + 8 LLM providers — from local Whisper & Ollama to cloud Deepgram & OpenAI
- Dual-party transcription — captures mic ("You") and system audio ("Them") simultaneously
- Real-time AI copilot — get streaming answers, follow-up suggestions, and meeting recaps from 8 LLM providers
- Local RAG pipeline — index your own documents (PDF, DOCX, TXT, MD) for context-aware AI responses
- 10 STT providers — Web Speech API, Deepgram, Groq, Whisper, ONNX Runtime, and more
- Always-on-top overlay — compact, transparent floating window visible only to you
- Bookmarks & action items — pin key moments and auto-extract tasks
- Speaker labeling — identify and name each speaker in the transcript
- Multi-language translation — real-time translation via 5 providers (100+ languages)
- Audio recording & playback — record meetings as WAV, replay with synced transcript
- Meeting scenarios — pre-configured templates for interviews, lectures, and team meetings
- Download the latest release
- Configure your STT and LLM providers (or use free local models)
- Start any meeting — NexQ captures system audio automatically
Getting Started Guide | All User Guides
| NexQ | Otter.ai | Granola | Krisp | |
|---|---|---|---|---|
| Price | Free | $8+/mo | $18/mo | $16/mo |
| 100% Local | Yes | No | Partial | Partial |
| Open Source | Yes | No | No | No |
| No Bot Joins | Yes | No | Yes | Yes |
| STT Providers | 10 | 1 | 1 | 1 |
| LLM Providers | 8 | 1 | 1 | 1 |
| Local LLM | Yes | No | No | No |
| RAG / Doc Context | Yes | No | No | No |
| Live Interview | Lecture Mode | Past Meeting Review |
|---|---|---|
 |
 |
 |
| Layer | Technology |
|---|---|
| Desktop | Tauri 2 (Rust + WebView2) |
| Frontend | React 18, TypeScript 5.5, Vite 6 |
| State | Zustand 4.5 |
| Styling | Tailwind CSS 3.4, shadcn/ui |
| Audio | cpal, WASAPI (Windows loopback) |
| STT | whisper-rs, ONNX Runtime, Deepgram, Groq, Web Speech API |
| LLM | OpenAI, Anthropic, Groq, Ollama, LM Studio, Gemini |
| Database | SQLite (rusqlite) |
# Clone the repository
git clone https://github.com/VahidAlizadeh/NexQ.git
cd NexQ
# Install frontend dependencies
npm install
# Run in development mode (launches Rust backend + React frontend)
npx tauri dev
# Build production installer
npx tauri buildnpm run dev # Vite dev server only (port 5173)
npm run build # TypeScript check + Vite production buildWhen you first run NexQ, Windows SmartScreen may display a warning. This is normal for open-source applications that are not code-signed. To proceed:
- Click "More info"
- Click "Run anyway"
Code signing certificates are expensive and not feasible for most open-source projects. The application is safe to run — you can verify by building from source.
Contributions welcome! See CONTRIBUTING.md for guidelines.
MIT License — free forever.
- Tauri — desktop application framework
- React — user interface library
- whisper-rs — Rust bindings for OpenAI Whisper
- Deepgram — speech-to-text API
- shadcn/ui — UI component library