Meetily: on‑device notetaker
- Zackriya Solutions’ Meetily has emerged as a fast-growing open-source meeting assistant that records, transcribes, and summarizes calls on local hardware. - The GitHub project shows about 11,300 stars, 1,100 forks, and a latest community release, v0.2.1, published on December 31, 2025. - Meetily is part of a broader push toward local-first workplace AI, trading cloud convenience for tighter control over sensitive meeting data (github.com) (meetily.ai).
Meeting note-takers usually ship audio to a vendor’s servers. Meetily is pitching the opposite: record, transcribe, and summarize meetings on your own machine. (github.com) The project is published by Zackriya Solutions on GitHub, where the repository showed roughly 11,300 stars and 1,100 forks when checked on April 25, 2026. The README says the Community Edition is free and open source. (github.com 1) (github.com 2) Meetily says it works across Zoom, Google Meet, Microsoft Teams, and other call platforms by capturing system audio and microphone input rather than building a deep integration into one conferencing app. Its website lists macOS, Windows, and Linux support, with Linux available from source. (dev.to) (meetily.ai) (github.com) The basic idea is simple: speech recognition turns conversation into text, then a language model turns that transcript into minutes, action items, and summaries. Meetily’s materials say the transcription runs locally and summaries can use local models, while newer releases also added support for OpenAI-compatible endpoints. (github.com 1) (github.com 2) That setup targets a specific problem in office software: cloud meeting bots can be convenient, but they also move sensitive discussions onto third-party infrastructure. Meetily’s pitch is that legal, health, defense, and enterprise users may want transcripts and summaries without handing raw audio to an outside service. (github.com) The tradeoff is that local-first software shifts more work to the user’s device and setup. Meetily’s own release notes describe model downloads, onboarding choices, and configuration around Whisper models, Ollama, and custom OpenAI-compatible providers. (github.com) (dev.to) The project has also been moving quickly. GitHub’s releases page shows v0.2.0 on December 31, 2025, followed by v0.2.1 the same day, with changes including automatic updates, improved onboarding, model fetching, and bug fixes for Windows audio selection and Whisper metadata. (github.com) Meetily’s site now pairs the open-source Community Edition with a paid Pro offering, including claims around custom summary templates, exports, and compliance-oriented features. That puts it in a familiar open-core lane: free local software for individuals, paid tooling for teams that want more polish. (github.com) (meetily.ai) For users comparing it with Otter, Fireflies, or Granola, the real choice is less about note-taking than where the computing happens. Meetily is betting that, for some meetings, the safest server is no server at all. (meetily.ai) (github.com)