Amanu

Records and transcribes online meetings. Automatically.

Amanu is a free, open-source meeting recorder for macOS and Windows. It works with Zoom, Google Meet, Telegram, WhatsApp, and other call apps without sending a bot into the meeting. It starts and stops recording on its own, separates speakers, writes a detailed summary, and keeps the complete record in an ordinary folder on your computer.

The current public downloads are macOS 0.6.5 and Windows 0.6.6.

Release builds read their version from the root VERSION file. The platforms can be published at different times; release tags and update feeds remain separate.

Recording window on macOS.

What Amanu does

• Records meetings automatically. Amanu notices when a call app is using

the microphone, adds optional calendar context on macOS, and stops when the call ends. Manual controls are always there too.

• Produces a speaker-attributed transcript. Your microphone and the other

side of the call remain distinct, with diarization inside each side when the transcription engine supports it.

• Puts names to voices. Amanu uses evidence in the transcript and, on

macOS, optional calendar participants. It accepts a name only when confidence is high. You can manually correct the rest.

• Writes a detailed summary. The result covers the topic, key points,

decisions, action items, and open questions. Amanu uses the models you choose instead of imposing a budget model of its own.

• Shows its work. The macOS status window, menu bar, and Dock icon, or the

Windows recording window and system tray, make it clear when a recording is running. An optional live transcript streams locally while you speak, independently of the engine chosen for the final transcript.

• Keeps one folder per meeting. Audio, transcript, speaker names, summary,

metadata, and processing logs are ordinary files that you own and can give to other tools.

Local when you want it, powerful when you need it

Recording always happens on your computer. Amanu can also transcribe and summarize a meeting without sending its contents anywhere:

• Parakeet, Whisper, and GigaAM provide local transcription. macOS local

transcription requires Apple Silicon; the Windows app bundles its x64 runtime and downloads the selected model from Settings.

• The optional live transcript runs locally.
• Ollama can write summaries locally when its Base URL is localhost/loopback.

Cloud models are available when quality or convenience matters more than staying entirely offline. On both platforms, AssemblyAI, OpenAI, and ElevenLabs