Comparison

Earshot vs Otter

Otter more or less invented the AI-notetaker category, and its real-time transcription is still excellent. The trade it asks for is your data: meetings transcribed and stored in Otter's cloud, used by default to improve Otter's models. Earshot makes the opposite trade — everything local, open source, your keys. Here's the comparison, fairly, as of July 2026.

The short answer

Pick Otter if you want live captions during the meeting, work across many devices, and are happy with cloud storage. Pick Earshot if you're on Windows and want your recordings and transcripts kept on your own machine, with the code open to inspection and the AI providers your choice.

At a glance

 EarshotOtter
Bot joins the callOptional (Otter Notetaker), plus bot-free desktop/extension
Where your data livesOtter's cloud (AWS, US)
Real-time transcript during the callYes, live
Open sourceNo
Self-host / own AI keysNo
Trains its own AI on your dataDe-identified, on by default (opt-out)
PlatformsWeb, Windows, Mac, iOS, Android, extension
Free tier300 min/mo, 30 min per conversation
Paid tierPro ~US$8.33/user/mo annual (US$16.99 monthly)

Verified against Otter's official pages in July 2026; Otter revises pricing periodically, so check its pricing page for current figures.

Where Otter wins, honestly

  • Live transcription. Otter shows the transcript as the meeting happens, with live speaker identification — genuinely useful for accessibility and for glancing back mid-call. Earshot's transcript and notes arrive after the meeting.
  • Everywhere. Web, Windows, Mac, iPhone, Android, a Chrome extension — Otter meets you on any device, and its mobile app covers in-person meetings.
  • Maturity. Years of polish, integrations and an AI chat over your whole meeting history, hosted for you with no setup at all.

Where Earshot wins

  • Your meetings never build up in someone else's cloud. Otter's model is storage: your conversations accumulate in your Otter account, on US-hosted infrastructure. Earshot stores recordings and notes on your PC — there's no server-side library of your working life.
  • No default training on your conversations. Otter uses de-identified customer data to improve its models unless you opt out. Earshot never trains on your content — there's nothing to opt out of.
  • Open source. Earshot's app is fully auditable on GitHub (MIT). Otter is a closed proprietary service.
  • Your AI, your rules. Self-hosted Earshot runs on the providers you pick — local Whisper, Groq, Deepgram, Anthropic, or a fully local model — so cost, quality and privacy are dials you control. And the free tier is the whole app, not 300 minutes.
  • Predictable pricing. One flat US$9/month for Plus (40 hours of transcription — roughly two hours every working day), or US$90/year. No per-seat maths, no 30-minute-per-conversation ceiling.

Which one should you pick?

If live captions during the call are the feature you actually use, Otter keeps that crown and Earshot won't pretend otherwise. For everything after the call — clean notes, action items you approve, answers with receipts — Earshot does the same job while your data stays yours. If that's the trade you want, download it free or start the 7-day Plus trial. Also see Earshot vs Granola, Fireflies and Fathom, or the full comparison.

Notes without the cloud archive.

Earshot records locally on Windows and keeps every meeting on your machine. Open source, free self-hosted, or Plus at $9/month.