Private AI meeting recorder — on-device
Record. Transcribe. Summarize. All on your phone.
Meeting Transcriber records any meeting or call, transcribes it, and delivers an AI summary with action items — 100% on-device. Your audio never leaves your phone.
100% on-device · your audio never leaves your phone
- 100% on-device
- No cloud uploads
- iOS + Android
100% on-device
Unlike Otter, Fireflies, and every cloud transcription tool, all processing happens on your device. GDPR/CCPA-friendly by construction.
Phone-native, in-person
Granola is Mac-only; Limitless requires a $99 pendant. We own the meeting that happens away from a laptop.
Fair $1/mo
Free tier with a daily allowance of premium actions. Pro at $1/month or $10/year — no ads, no watermarks.
How it works
Three steps from meeting to action items
Tap to record
One tap starts recording via the microphone. Works in-person, on a call, in the field — anywhere you have your phone.
Transcribe on-device
When you stop, the app transcribes locally using an on-device speech model. No audio leaves your phone.
AI summary + action items
An on-device LLM produces a concise summary, key points, and extracted action items — ready to share.
See it in action
From recording to summary, all on your phone
What you get
Everything a recorder needs — nothing it doesn't
On-device transcription
Speech-to-text runs entirely on your device. No round-trip to a server, no waiting, no audio uploads.
AI summary + action items
An on-device LLM distills the full transcript into a summary, key points, and a clear action-item list.
Transcript playback sync
Tap any line in the transcript to jump the audio to that exact moment — perfect for verifying what was actually said.
Full history + search
Every recording, transcript, and summary is stored locally and searchable. No account required.
One-tap export
Share as plain text, Markdown, or copy to clipboard — free, no watermark. PDF export is a premium feature.
10 languages
Transcription and AI summary in English, Spanish, French, German, Italian, Portuguese, Russian, Chinese, Japanese, and Korean.
Privacy by design
Your meetings are nobody's business but yours
- Audio never uploaded — processing is entirely on your device, zero server round-trips for the core loop.
- On-device speech-to-text: the transcript exists only in your app's local storage.
- On-device LLM summary: the model weights are downloaded once and run fully offline thereafter.
- Model downloads are the only network touch — one-time, content-free (no audio, no transcript text).
- Anonymous Firebase Analytics + Crashlytics for stability only — no audio content, no PII.
- RevenueCat for subscriptions — purchase token only, never audio or transcript.
Simple pricing
Real value on free. Unlimited on Pro.
On-device inference costs nothing to serve, so nothing is hard-locked. Free users get real value every day. Pro removes the daily cap.
Free
$0
forever
Unlimited recording, transcription, and key-points summary. Plus 3 premium actions per day.
- Unlimited microphone recording
- On-device transcription
- Key-points summary (unlimited)
- Full local history + search
- Text + Markdown export
- 3 premium actions per day (full summary, action items, PDF)
Pro
$1
per month
All daily caps removed. Unlimited premium actions — AI summaries, action items, follow-ups, PDF export.
- Everything in Free
- Unlimited full AI summary + action items
- Unlimited follow-up draft generation
- Unlimited PDF export
- Unlimited system/call-audio capture
- Priority on-device model tier
Get 2 months free with the annual plan — $10/year vs. $12/year monthly. No auto-renewing surprises.
Questions
Good to know
- Does my audio ever go to the cloud?
- No. Audio processing happens entirely on your device. The only network requests the core loop makes are a one-time model-weights download (cached locally) and telemetry/auth pings (Firebase, RevenueCat) that contain no audio or transcript content.
- Which languages are supported for transcription and summary?
- English, Spanish, French, German, Italian, Portuguese, Russian, Chinese (Simplified), Japanese, and Korean. Both transcription and the AI summary run in all ten languages on both iOS and Android.
- Is it available on both iOS and Android?
- Yes — Meeting Transcriber is built with Kotlin Multiplatform and ships on both platforms, with on-device engines optimized for each.
- What is the daily allowance?
- Free users get 3 premium actions per day (full AI summary + action items, follow-up draft, PDF export, system-audio capture). The counter resets each calendar day. The basics — unlimited recording, transcription, key-points summary, history, and text/Markdown export — are always free with no daily limit.
- What does Pro include?
- Pro removes every daily cap. Unlimited full AI summaries, action-item extraction, follow-up drafts, PDF exports, and system/call-audio capture. The only difference between Free and Pro is that Pro lifts the per-day limit on premium functions.
- Does system/call-audio capture work on both platforms?
- System-audio capture (recording Zoom, phone calls, or device audio) is a premium function. Microphone capture (in-person, phone-on-speaker) is always free and unlimited. System-audio support depends on platform permissions.
- Is the app GDPR/CCPA compliant?
- By construction. Audio and transcripts never leave your device, so there is no cloud data processor to contract with for the core function. Anonymous Firebase Analytics and Crashlytics are the only data touchpoints, and they carry no audio or transcript content.