{"slug": "show-hn-minute-an-offline-meeting-notetaker-using-local-llms", "title": "Show HN: Minute – An Offline Meeting Notetaker Using Local LLMs", "summary": "A developer released Minute 1.14.0, a free, MIT-licensed, Apple-signed and notarized macOS app that records, transcribes and summarizes meetings entirely on the user's own machine using local Whisper models and a local language model, with no account and no bot joining the call. Minute requires macOS 11 or later, needs macOS 13 or later plus Screen Recording permission to capture system audio, and needs 8 GB of memory for smaller models and 16 GB for larger ones; the app's network ledger states audio, transcripts and notes are never sent, with only a one-time model download and an optional GitHub update check. The developer positions Minute against Granola Business at $14 per user per month and Otter Pro at $16.99 per user per month, noting Otter sends a bot into Zoom, Teams and Meet calls while Minute does not.", "body_md": "### Live transcript as people speak\n\nWhisper runs on your Mac and keeps pace with the conversation. Click a timestamp to play that moment.\n\nFree, offline AI note taker for macOS\n\nMinute records your call, transcribes it, and writes the summary on your own machine. No account. No bot in the call. Open source.\n\nVersion 1.14.0 · Signed and notarized by Apple · macOS 11 or later\n\nWorks with the call apps you already use. No bot joins.\n\nMinute notices the meeting, offers to record, and shows the transcript as people speak. When the call ends, the notes are already written.\n\nMinute records your microphone and, on macOS 13 or later, the call's system audio. That is how it hears the other people. Nobody sees a new participant, and nobody has to let one in.\n\nSystem audio needs the Screen Recording permission. Without it, Minute records your microphone only. Meeting detection covers Zoom, Microsoft Teams, Webex, Slack, FaceTime, Discord, and browser calls such as Google Meet.\n\nThat's the whole list. Turn off Wi-Fi and everything still works.\n\nThis website counts visits with [Plausible](https://plausible.io/data-policy), which uses no cookies. The Minute app has no analytics at all.\n\n| Network ledger | What happens | \n|---|---|\n| Your audio | Never sent | \n| Your transcripts and notes | Never sent | \n| Analytics or telemetry in the app | None | \n| Model download | Once, on first use | \n| Update check against GitHub | Optional, can be switched off | \n\nEvery step runs on your Mac: recording, transcription, the summary, and your questions.\n\nMinute works out who said what, and can remember the voices you name.\n\nWritten by a local language model when the call ends.\n\nAsk any meeting a question. Each answer cites the moment it was said, and a click on the time jumps there.\n\nMinute offers to record when a call starts and to stop when it ends. It is off by default.\n\nSave each summary to your vault, or export any note as Markdown.\n\nWhisper in 99 languages, and a summary model that fits your Mac's memory.\n\nGranola Business is $14 per user per month. Otter Pro lists at $16.99.\n\nMinute MIT licensed\n\n**$0** No account. No trial.\n\n| Feature | Minute | Granola | Otter | \n|---|---|---|---|\n| Price | Free | Basic $0. Business $14 per user per month. | Basic free. Pro $16.99 per user per month. | \n| Free plan limits | None | 30 days of history | 300 minutes a month, 30 per conversation | \n| Transcription | On your Mac | In the cloud | In Otter's cloud | \n| A bot joins the call | No | No | Yes, on Zoom, Teams, and Meet | \n\nChecked on 10 November 2026 against [granola.ai/pricing](https://www.granola.ai/pricing), [granola.ai/security](https://www.granola.ai/security), and [otter.ai/pricing](https://otter.ai/pricing).\n\n[Minute as a Granola alternative](granola-alternative/)\n[Minute as an Otter alternative](otter-alternative/)\n\nYes. Minute is free and open source, and it needs no account.\n\nYes. After you download the models once, Minute records, transcribes, summarizes, and answers questions with Wi-Fi turned off. It goes online only to download a model, to list models, and to check for an update. You can turn off the update check in Settings.\n\nNo. Minute records on your Mac. It captures your microphone and, on macOS 13 or later, the system audio, so it hears the other people on the call without joining it. Meeting detection covers Zoom, Teams, Webex, Slack, FaceTime, Discord, and calls in a browser, such as Google Meet.\n\nMinute needs macOS 11 or later. System audio needs macOS 13 or later and the Screen Recording permission. To notice that a call ended while the meeting app stays open, Minute needs macOS 14.2 or later. Apple silicon is the tested target. The Intel build is not yet tested on real Intel hardware, and the large Whisper models need Apple silicon. The smaller models need 8 GB of memory, the larger ones 16 GB.\n\nYes. Choose your vault folder in Settings, and Minute writes a Markdown file each time a summary finishes. It can add the date, title, duration, and speakers as properties. Without Obsidian, export any note as Markdown.\n\nYes. The [code is on GitHub](https://github.com/mraza007/minute) under the MIT license, and each release is signed and notarized by Apple.\n\nMinute is a Rust app with whisper.cpp and llama.cpp compiled in, running on Metal. It has no Electron and no Python runtime. Notes are plain text and WAV files in a folder you own.\n\nGranola and Otter transcribe in the cloud, and Otter's assistant joins Zoom, Teams, and Google Meet calls. Minute writes the transcript and the summary on your Mac, with no bot. Read how Minute compares as a [Granola alternative](granola-alternative/) and as an [Otter alternative](otter-alternative/), including when each of those apps is the better choice.\n\nMinute is the free, offline AI note taker for macOS.", "url": "https://wpnews.pro/news/show-hn-minute-an-offline-meeting-notetaker-using-local-llms", "canonical_source": "https://minutenotes.app/", "published_at": "2026-11-15 21:06:26+00:00", "updated_at": "2026-11-15 21:22:32.359473+00:00", "lang": "en", "topics": ["ai-tools", "artificial-intelligence", "large-language-models", "ai-products"], "entities": ["Minute", "Whisper", "Granola", "Otter", "Apple", "Zoom", "Microsoft Teams", "Google Meet"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/show-hn-minute-an-offline-meeting-notetaker-using-local-llms", "markdown": "https://wpnews.pro/news/show-hn-minute-an-offline-meeting-notetaker-using-local-llms.md", "text": "https://wpnews.pro/news/show-hn-minute-an-offline-meeting-notetaker-using-local-llms.txt", "jsonld": "https://wpnews.pro/news/show-hn-minute-an-offline-meeting-notetaker-using-local-llms.jsonld"}}