Open-source meeting transcription
Apps that record a call or a file, produce a transcript, and in some cases a summary, without sending the audio to a vendor. The summary step often uses a language model you choose, local or hosted.
A meeting assistant under MIT that records your microphone and system audio together, transcribes live on your machine with Whisper or Parakeet, and summarises with a language model you choose.
A desktop app under MIT that transcribes and translates audio and video files, YouTube links and a live microphone offline with Whisper, with speaker identification and SRT, VTT and TXT export.
An offline transcription app under MIT that records system audio and the microphone, supports Whisper, Parakeet and Nemotron models, labels speakers, and exports SRT, VTT, TXT, HTML, PDF, JSON and DOCX.
Questions people ask
What is the best open-source meeting transcription tool?
For a live call on your own machine, Meetily (MIT) records microphone and system audio together, transcribes with Whisper or Parakeet, and summarises with a model you choose. OpenWhispr (MIT) detects Zoom, Teams and FaceTime calls and labels speakers on device. For recordings and files, Buzz and Vibe (MIT) transcribe, label speakers and export documents.
Can open-source tools transcribe a Zoom or Teams meeting?
Yes, from the machine you are on. Meetily and Vibe capture system audio, which is how they hear the other side of a call, and OpenWhispr detects when a Zoom, Teams or FaceTime call starts. None of them joins the call as a separate bot participant the way Otter and Fireflies do.
Which open-source meeting tools label who said what?
OpenWhispr labels speakers live on device with voice fingerprints that carry across meetings, though not on Intel Macs. Buzz identifies speakers in transcribed media. Vibe lists speaker diarization. Meetily's community edition README does not list it; it appears in the paid PRO edition as coming soon.
Do open-source meeting tools keep the audio private?
Meetily, Buzz and Vibe transcribe entirely on your machine. Meetily's README says no data leaves your computer when the summary model is local too. Vibe's summary through the Claude API is a cloud call; its Ollama path is not. OpenWhispr is local when you choose local models.
Paying for an AI tool you would rather own?
Reveneau sets up the replacement on your own servers, connects it to your product, and builds the software around it.