Wispr Flow has officially stepped out of the dictation lane and into the crowded AI meeting-notetaker market. On August 5, 2026, the company launched Notetaker, a Mac-based tool that captures system audio directly from your device — no meeting bot required
. Here's what it does, how it stacks up against the competition, and what the updated privacy policies mean for users.
Notetaker Features: What You Get
Notetaker is built around a core idea: capture everything said in a meeting and turn it into a useful, searchable artifact — without anyone knowing a bot is present. It offers a slate of AI-powered capabilities:
- Live transcription – Speech appears in real time during the meeting, and after the call ends Wispr produces a refined transcript with speaker names .
- AI catch-up ("What did I miss?") – A dedicated button that summarizes what was said before a late joiner arrived, surfaced in the Chat tab during recording .
- Cross-meeting AI querying – Users can ask questions across their entire meeting history and get AI-generated answers .
- Real speaker names – The final transcript labels who said what, rather than "Speaker 1" and "Speaker 2" .
- Clean summaries – Automatically extracted decisions, action items, and structured notes .
- Personal dictionaries – Notetaker pulls from the user's existing Wispr Flow dictionary (acronyms, product names, custom terminology) to improve accuracy .
- Standalone notepad – Users can type notes during the call in a dedicated window, with the final page combining user notes, AI summary, and transcript .
- Auto-separate meetings – Back-to-back sessions in the same room automatically start fresh recordings so conversations don't merge .
The tool also integrates with calendars, Gmail, and Slack to pull context before and during meetings .
Wispr Notetaker vs. Granola: Key Differences
Both Wispr Notetaker and Granola use device-level audio capture — no meeting bot joins the call
. But they differ significantly in execution: