Type: GitHub Repository
Original Link: https://github.com/FrigadeHQ/yogurt
Publication Date: 2026-11-20
Summary #
Introduction #
Imagine being in an important meeting: your team discusses next quarter’s strategy, crucial decisions emerge, specific tasks are assigned. Meanwhile, you’re trying to take notes, but you lose the thread of the conversation. Or worse, you invite a third-party bot to record, and your data ends up on remote servers you don’t control. Yogurt solves this dilemma elegantly: an AI assistant that records, transcribes, and summarizes your meetings, but everything stays on your computer. No cloud, no mandatory accounts, no privacy compromises.
With over 2600 stars on GitHub, Yogurt has established itself as the open-source alternative to Granola that many developers and professionals were waiting for. It’s available on Mac, Windows, and Linux, and represents a paradigm shift in how we think about collaboration tools: powerful, intelligent, but radically private.
What It Does #
Yogurt transforms your computer into an intelligent meeting assistant. When you start a call on Zoom, Teams, or Google Meet, Yogurt captures the audio from your microphone and the conversation, converts it to text in real-time using Whisper, and then automatically generates a meeting summary. But it’s not just transcription: the system identifies who’s speaking (on Mac, Windows, and Linux), supports 99 languages, and can even follow multilingual conversations where participants switch languages mid-discussion.
What makes Yogurt different is that all of this happens locally. While you’re talking, your audio is processed by the AI model directly on your machine. There’s no bot joining the meeting, no remote server recording your data. You can add personal notes during the call, search through your past meetings by title, transcript, or folder, and export everything as Markdown. If you wish, Yogurt can also act as a local MCP server, allowing other applications on your computer to access your notes, transcriptions, and calendar details.
Why It’s Amazing #
The “wow” factor of Yogurt is the combination of power and privacy in a single desktop application. It’s not just a simple recorder that saves audio files: it’s a complete natural language processing system that works offline.
Radical privacy as foundation: This isn’t a compromise or an added feature, it’s the architecture itself of the project. Your audio data, transcriptions, and AI processing remain on your computer. There’s no cloud synchronization, no hidden telemetry, no mandatory account. If you want to use Yogurt, you download it, install it, and start right away. This “local-first” philosophy is particularly important for those working with sensitive information: legal consultants, doctors, financial professionals, or simply anyone who prefers their data to remain private.
Contextual intelligence in real-time: Yogurt doesn’t just passively transcribe. While you talk, the system understands the context of the conversation and generates intelligent summaries that capture not only what was said, but also the decisions made, next steps, and who’s responsible for what. You can ask questions to your notes and the system answers you with references to specific sources, allowing you to find exactly where information was mentioned.
Multilingual and adaptive: In an increasingly global world, Yogurt handles conversations in 99 different languages. If your meeting includes people speaking Italian, English, and Spanish, the system follows each language separately and maintains transcription coherence. It’s not an approximation: it’s true support for international work.
No bot in the meeting: Unlike many recording services, Yogurt doesn’t add an artificial participant to your call. It captures audio from your computer without other participants knowing (of course, always respect local recording laws). This means more natural meetings, without the distraction of a bot joining and disconnecting.
How to Try It #
Yogurt is available for direct download from GitHub Releases, or on Mac you can use Homebrew with a simple command: brew install --cask frigadehq/yogurt/yogurt. Requirements are minimal: at least 8 GB of RAM and a recent operating system (macOS 14.4+ for Apple Silicon, Windows 10 2004+ for Windows, Ubuntu 22.04+ for Linux).
Installation is immediate. On Mac, extract the DMG and drag Yogurt to the Applications folder. On Windows and Linux, follow standard instructions for your operating system. Once launched, Yogurt offers to automatically record when it detects a supported call. You can start recording manually at any time, add notes during the meeting, and when it’s done the transcript and summary are generated automatically.
Complete documentation is available in the GitHub repository and on the official yogurt.so website. There’s no one-click online demo, but the project is designed to be intuitive: download, install, record. If you want to contribute or customize the code, the repository is completely open-source under MIT license, with detailed instructions for building locally.
Final Thoughts #
Yogurt represents an important shift in how we think about productivity tools. In an era where personal data has become a precious commodity, this project demonstrates that it’s possible to build intelligent and powerful tools without sacrificing privacy. It’s a concrete alternative to proprietary services, maintained by the community and available to everyone for free.
For those working in distributed teams, for those with confidentiality needs, or simply for those who prefer to control their own data, Yogurt is a tool worth trying. With over 2600 stars and an active community, the project continues to evolve. If you believe meetings should be yours and yours alone, Yogurt is exactly what you were looking for.
Use Cases #
- Private AI Stack: Integration in proprietary pipelines
- Client Solutions: Implementation for client projects
- Development Acceleration: Reduction of project time-to-market
Resources #
Original Links #
- GitHub - FrigadeHQ/yogurt: Private AI meeting notes that stay on your computer. A free… - Original link
Article reported and selected by the Human Technology eXcellence team processed through artificial intelligence (in this case with LLM HTX-EU-Claude-Haiku-4.5) on 2026-11-20 10:44 Original source: https://github.com/FrigadeHQ/yogurt
Related Articles #
- GitHub - block/buzz: A hive mind communication platform - Rust, Open Source, AI Agent
- GitHub - yichuan-w/LEANN: RAG on Everything with LEANN. Enjoy 97% storage savings while running a fast, accurate, and 100% private RAG application on your personal device. - Python, Open Source
- GitHub - jundot/omlx: LLM inference server with continuous batching and SSD caching for Apple Silicon — managed from the Mac - Machine Learning, LLM, Python