Skip to content
Get packs

Guide · Open source

How to use Meetily: install, transcribe and summarize meetings

Updated · written with AI from the project's own documentation and checked against our catalogue data

Meetily is a free, MIT-licensed desktop app that records your meetings, turns the audio into a live transcript and can generate summaries. This guide shows you how to install it on macOS or Windows, what it does with your audio, and how to choose a model and summary provider. By the end you will know how to run a meeting through Meetily and where its data is stored.

MeetilyZackriya-Solutions/meetily · ★ 32.4k · MIT

Quick start

  1. 1Download the app. Get the Windows x64-setup.exe or the macOS .dmg from the Releases page on GitHub.
  2. 2Install and open Meetily. Run the Windows installer or drag Meetily into Applications on macOS, then open it.
  3. 3Choose a transcription model. Select Whisper or Parakeet for on-device transcription.
  4. 4Select audio devices. Pick your microphone and system audio so both sides of the meeting are captured.
  5. 5Record or import audio. Start a live meeting recording, or import an existing audio file using the Beta import feature.
  6. 6Configure a summary provider. Use Ollama for local summaries, or enter Claude, Groq, OpenRouter, OpenAI or your own OpenAI-compatible endpoint.

Install Meetily on Windows or macOS

Meetily is distributed as a downloadable app for macOS and Windows. Get the latest build from the project's Releases page on GitHub.

On Windows, download the latest x64-setup.exe and run the installer. On macOS, download the meetily .dmg file, open it, and drag Meetily into your Applications folder, then open it from there.

  • Windows: the packaged installer uses a Vulkan-enabled Whisper build and needs an AVX2-capable x64 CPU. AVX-512 is not required.
  • macOS: the package listed in the README is built for Apple silicon (aarch64).

Understand what runs on your device and what goes to a provider

Transcription runs on your own device using Whisper or Parakeet models, so audio and transcripts are meant to stay local. The README says transcription models, recordings and transcripts are stored on your machine.

Summaries are different. You choose the AI provider. Ollama runs a model locally and is the option the README recommends. You can also use Claude, Groq, OpenRouter, OpenAI, or your own OpenAI-compatible endpoint. Cloud providers receive the text you send them for summarizing.

Record and transcribe a meeting

Meetily captures your microphone and system audio at the same time, with ducking and clipping prevention, so both sides of a call can appear in the transcript. The transcript is generated live as the meeting happens.

You can also import an existing audio file to generate a transcript. The README marks this Import & Enhance feature as Beta. Enhance re-transcribes a recorded meeting with a different model or language.

  • Pick the audio devices for your microphone and system output in the app settings.
  • Choose a Whisper or Parakeet model for transcription.

Set up AI summaries

Summaries are generated from the transcript using the provider you select. For fully local summaries, run Ollama and select it as the provider. For a team setup, point Meetily at your own OpenAI-compatible endpoint through the custom endpoint setting.

The README does not list specific model names or required hardware for summaries, so check the settings screen and your provider's documentation before relying on a particular model.

Build from source on Linux

The README does not provide a packaged Linux installer. Linux users build from source. Clone the repository, install the frontend dependencies and run the GPU build script from the frontend folder.

The README points to docs/building_in_linux.md and docs/BUILDING.md for full instructions. The GPU script's acceleration options depend on your hardware, and the README says to use the configuration appropriate for your platform.

git clone https://github.com/Zackriya-Solutions/meeting-minutes cd meeting-minutes/frontend pnpm install --frozen-lockfile ./build-gpu.sh

Know the limits before you rely on it

Speaker diarization is planned for the paid Meetily PRO edition, and the README does not clearly state its status in the free build. Do not plan around it being available in the free version.

GPU acceleration depends on platform and build. macOS packages enable Metal and CoreML automatically. Windows packages use Vulkan. CUDA on Windows requires a source build with a compatible NVIDIA toolchain.

  • Linux needs a source build.
  • Diarization status in the free build is not clearly documented.

What people use it for

  • Taking private notes for sensitive client or legal conversations without uploading audio to a cloud note taker.
  • Re-transcribing a recorded meeting with a different model or language using Enhance.
  • Generating summaries through a company's own OpenAI-compatible AI endpoint.
  • Transcribing existing audio files you already have from earlier meetings.

Frequently asked questions

Is Meetily free?

Yes. The README describes Meetily as open source and free to use, under the MIT licence. Meetily PRO is a separate paid edition with extra features such as enhanced accuracy and advanced exports.

Does Meetily work on Linux?

The README says Linux requires building from source. It does not offer a packaged installer for Linux, so you clone the repository and run the build steps.

Does Meetily send my recordings to the cloud?

The README says transcription runs on your device and recordings and transcripts are stored locally. Summaries can go to a cloud provider only if you choose one, such as Claude, Groq, OpenRouter or OpenAI.

Can I use Meetily fully offline?

Transcription is designed to run locally with Whisper or Parakeet. For offline summaries, the README names Ollama as the local option, so use it as your summary provider.

Does Meetily support speaker diarization?

The README says speaker diarization is planned for the PRO edition in mid-June. The status for the free build is not clearly stated, so check the project's current documentation before depending on it.

What is the alternative to Meetily?

Comparable projects listed for this tool include Whisper.cpp, Scriberr and Hyprnote. Whisper.cpp is a transcription engine, while Scriberr and Hyprnote are meeting-oriented tools. Choose based on whether you need a full desktop app or a lower-level component.

New agent skills, repos and AI tools — every week

The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.

Our verdict

Meetily suits people on macOS or Windows who want meeting notes that stay on their own machine and a free, MIT-licensed app they can control. It is less suitable for Linux users who want a packaged installer, or for anyone who needs cloud-only convenience and does not want to manage local models or providers.

Install options & downloads