| How meetings are captured |
You manually start microphone-only local capture for meetings you can hear. No bot joins, but remote system audio is not captured. |
You manually start botless capture of microphone and system audio on the device. No bot joins, and there is no calendar auto-join or auto-record. |
| Where transcription runs |
Meeting transcription always runs on-device with local engines; it stays local even if you use opt-in Jot Cloud for dictation. |
Audio is captured locally, then streamed to cloud transcription providers Deepgram or AssemblyAI. Cloud AI providers create enhanced notes. |
| Works offline |
Yes, after the local model download. |
No offline mode documented; transcription and AI note enhancement depend on cloud services. |
| AI summaries and action items |
None. WhisperJot provides a transcript to review and export. |
Blends the transcript with your typed notes into AI-enhanced summaries, with templates, live transcript view, and chat across meetings. Speaker labels depend on setup and often distinguish only Me and Them. |
| Price |
$99/year with no per-minute or storage caps; meetings can last up to six hours. |
Basic is free. Business is $14/user/month with monthly billing; annual billing is available through sales. Enterprise starts at $35/user/month. |
| Free tier |
No free tier; a 30-second browser demo is available. |
Basic provides unlimited meeting notes, but history access is limited to the last 30 days. Older notes remain stored and become accessible again after upgrading. |
| What happens to your audio |
Audio stays on your machine, with configurable retention that defaults to 90 days. Optional cloud library sync includes text, titles, and timestamps, never audio. |
Meeting audio is not retained after cloud transcription, so post-meeting playback is unavailable. Transcripts are encrypted in a US-hosted AWS environment. Granola trains on anonymized data by default for non-Enterprise users, with a Settings opt-out. |