Memory Notch
CLOUD SCRIBES
Memory Notch vs Fathom
A generous free tier for recorded calls, with a bot on the invite.
| Memory Notch | Fathom | |
|---|---|---|
| Where the audio goes | Nowhere | To Fathom's servers |
| Processing | On the Neural Engine | Cloud |
| Bot in the call | No | Yes, a participant joins the meeting |
| Speaker labels | Yes, on-device | Yes, in the cloud |
| Platform | macOS 26 | Web, macOS, Windows |
| Open source | No | No |
| Price | Not released yet | Free tier, Premium $20/month, Business $34/user/month |
Details checked September 2026 against Fathom's own documentation. Prices move; treat the figure as a starting point, not a quote.
What Fathom does well
- The free tier records without a time limit, which is rare.
- Fast, clean summaries and clip sharing.
- Straightforward CRM sync on the paid tiers.
What it costs you
- Everything is processed and stored on Fathom's infrastructure.
- Free-tier AI summaries are capped per month.
- A bot joins, so the recording is announced to the room.
Where Memory Notch is different
Memory Notch does the transcription and the speaker separation on your Mac's Neural Engine. There is no account, no telemetry and no API key, and a recording is never sent anywhere. The models download once, and after that a recording transcribes with the network off. Each one becomes a directory of plain Markdown in ~/memory, which means the AI tool you already use can read it by path: no MCP server, no export step, no API.
It is also a smaller thing than Fathom. One Mac, one person, no team workspace and no integrations. If you need a shared archive or a CRM, Fathom is the better tool and this page is not trying to talk you out of it.