Granola is an AI meeting notepad founded in 2023 by Chris Pedregal, who previously built and sold the education app Socratic to Google, and product designer Sam Stephenson. Its defining design choice is that no bot joins your calls: the desktop app captures system audio directly, so participants see nothing unusual and there is no awkward "recording assistant has joined" moment. Worth being precise about what that does and does not mean — capture is local, but transcription and summarization run in the cloud through third-party subprocessors, so this is a no-bot product rather than an on-device one.
The second design choice is that you keep typing. Granola expects you to jot rough shorthand during the meeting, then merges those notes with the full transcript into a structured writeup that follows your emphasis rather than a generic summary template. In practice this is why users describe it as the rare AI notetaker they keep using — the output looks like notes they would have written with more time.
Through 2026 the product has been expanding past note-taking into what the company calls an enterprise context layer: chat across your entire meeting archive, shared folders, an API, and an MCP server so other AI tools can query meeting history. Granola raised $125M at a $1.5B valuation in March 2026 led by Index Ventures, following a quarter in which revenue grew 250%, and has raised roughly $192M in total.
Worth being clear about the limits: Granola is an excellent capture-and-retrieval tool, not an autonomous agent. It does not yet execute follow-ups, update your CRM unprompted or run multi-step tasks — the company has signalled agentic features are coming but they are not shipped. Platform coverage has only just filled out, too: Android arrived on 1 July 2026, so the mobile apps are considerably newer and thinner than the macOS original.
Key Benefits
- Unobtrusive capture: Direct system-audio recording means no meeting bot, no participant friction and no separate join link.
- Notes that match your intent: Merging your shorthand with the transcript beats generic auto-summaries for accuracy of emphasis.
- Searchable institutional memory: Chat spans every meeting you've had, turning months of calls into a queryable knowledge base.
- Open to other tools: MCP and API access let Claude, internal agents or CRM automations pull meeting context directly.
- Fair pricing: $14/user/month for unlimited history and integrations is well below most enterprise meeting-intelligence tooling.
Use Cases
- Founder and executive meeting archive — Capture every investor, customer and team call, then ask questions across months of history instead of scrolling documents.
- Sales and CRM hygiene — Push structured call notes into HubSpot, Attio or Affinity so pipeline records stay current without manual write-ups.
- Consulting and client delivery — Use per-client templates and shared folders so the whole engagement team works from the same meeting record.
- Feeding downstream AI tools — Expose meeting context over MCP so a coding, research or ops agent can act with knowledge of what was actually decided.
