Meetily is a privacy-first AI meeting assistant built in Rust that performs real-time transcription, speaker diarization, and meeting summarization entirely on local machines without requiring cloud connectivity. The application runs on macOS and Windows, with Linux support available through source builds, and represents an open-source alternative to cloud-based meeting intelligence tools. The project emphasizes data sovereignty and compliance, positioning itself as a solution for enterprises and professionals handling sensitive information across defense, legal, healthcare, and executive sectors.
The core functionality centers on local-first processing using speech-to-text models including Parakeet and Whisper, with the repository documentation claiming 4x faster transcription speeds compared to standard Whisper implementations. Transcription happens in real-time as meetings occur, with the system capturing both microphone and system audio simultaneously through intelligent audio mixing that includes ducking and clipping prevention. The application supports hardware acceleration across different platforms, including Apple Silicon with Metal and CoreML on macOS, and NVIDIA CUDA or AMD/Intel Vulkan support on Windows and Linux systems.
Meeting summarization leverages multiple AI provider options, with Ollama recommended for fully local processing. The system also supports Claude, Groq, OpenRouter, and custom OpenAI-compatible endpoints, allowing organizations to integrate their preferred AI infrastructure. An import and enhancement feature enables users to process existing audio files or re-transcribe previously recorded meetings using different models or languages, all handled locally without external services.
The technical architecture uses Tauri as the application framework, combining a Rust backend for core logic with a Next.js frontend interface. The repository is licensed under MIT and distributed through GitHub releases with pre-built installers for Windows and macOS. The project maintains active development with features in beta status, including the import and enhancement functionality contributed by community members.
Meetily addresses specific privacy and compliance concerns, citing statistics including a 4.4 million dollar average cost per data breach according to IBM 2024 data and 5.88 billion euros in GDPR fines issued by 2025. The documentation emphasizes that all processing, model storage, recordings, and transcripts remain on the user's device with no data transmission to external servers. This approach eliminates vendor lock-in and provides complete control over sensitive meeting content.
The project offers both a free Community Edition and a paid Meetily PRO tier. The Community Edition remains permanently free and open source, while PRO targets users and teams requiring enhanced transcription accuracy, custom summary templates, advanced export options, auto-meeting detection, and self-hosted deployment capabilities. Speaker diarization functionality was planned for the PRO tier in mid-June. A promotional coupon code LAUNCH20 provided 20 percent discount on PRO subscriptions through the next Community Edition release cycle.
The repository includes comprehensive documentation covering system architecture, building from source, and installation procedures across supported platforms. Development requires Rust and Node.js installations for contributors. The project maintains community engagement through Discord channels, Reddit communities, and LinkedIn presence, with a website at meetily.ai serving as the primary information hub and product update subscription point.
