| How meetings are captured |
Microphone-only local capture for meetings you can hear: in-person meetings, interviews, lectures, and your side of calls. It does not capture system audio or the remote side of a video call, and no bot joins. |
An AI notetaker bot can join Zoom, Google Meet, or Teams through calendar auto-join. Mac and Windows desktop apps also provide botless local system-audio capture. |
| Where transcription runs |
Meeting transcription always runs on-device with the local engines, including for opt-in Jot Cloud dictation users; meeting audio never streams to the cloud. |
Cloud transcription on US-based AWS. Otter says live transcription requires a stable internet connection; recordings can continue locally and transcribe after reconnecting. |
| Works offline |
Yes, after the local model download. |
Not for live transcription. Recording can continue locally while offline, with transcription after the connection returns. |
| AI summaries and action items |
None. WhisperJot produces a transcript for you to review and export. |
Includes speaker identification, AI summaries, Otter AI Chat, action items, slide capture, and meeting and CRM integrations. |
| Price |
$99/year for meeting notes and system-wide dictation, with no per-minute or storage caps and a six-hour cap per meeting. |
Basic is free. Pro is $8.33/user/month annually or $16.99 monthly; Business is $19.99/user/month annually or $30 monthly; Enterprise is custom. |
| Free tier and limits |
No free tier; a 30-second browser demo is available. |
Basic includes 300 transcription minutes/month, 30 minutes/conversation, and three lifetime file imports. Pro includes 1,200 minutes/month and 90 minutes/conversation. Business has unlimited transcription with a four-hour conversation cap. |
| What happens to your audio |
Audio stays on your machine, with 90-day retention by default that you can change. Optional cloud library sync sends text, titles, and timestamps, never audio. |
Audio, screenshots, and transcripts are stored on US AWS as long as necessary. Otter's policy says it trains proprietary AI technology on de-identified audio and on transcriptions that may contain personal information. |