VoiceSumm
iPhone / Productivité
Voice notes + AI summary, running entirely on your device.
Most transcription apps either can't summarize, or they upload your audio (or transcript) to a server to do it. VoiceSumm does both — transcription AND summary — and your recordings never leave your device.
VoiceSumm runs BOTH steps on-device:
• Whisper transcribes your audio locally.
• A bundled Llama 3.2 3B language model reads the transcript and produces a structured summary — bullets, action items, decisions.
• Both models live inside the app. No network calls for transcription or summary — your recordings and transcripts never leave your device.
WHY THIS MATTERS
Most "private" transcription apps stop at transcription. The moment you want a summary, your audio (or transcript) leaves your device. VoiceSumm doesn't. The summary is generated by a 3-billion-parameter language model running entirely on your device. Your conversation never touches a server you don't own.
WHO IS THIS FOR
• Lawyers, paralegals, and consultants whose meetings carry attorney-client or NDA risk.
• Doctors, therapists, and counselors whose patient conversations belong to the patient, not a cloud vendor.
• Journalists protecting their sources, especially in jurisdictions where cloud AI is restricted.
• Founders, operators, and investors recording sensitive strategy and due-diligence calls.
• Anyone who wants the convenience of AI summary without giving away the audio that produced it.
HOW IT WORKS
1. Tap the mic button. The app records locally.
2. When you stop, Whisper transcribes the audio on your device.
3. Llama 3.2 3B reads the transcript on your device and produces a structured summary with key points, action items, and decisions.
4. Share via the system share sheet (Mail, Files, Messages, Notes) — or copy as Markdown.
NO CLOUD, BY DESIGN
VoiceSumm has no servers. No account to create, no login, no subscription, no analytics, no telemetry. Your audio and transcripts never reach anyone else's computer.
PRICING
$14.99 one-time. No subscription, no "Pro upgrade," no surprise auto-renewal. Pay once, own forever.
LANGUAGES
Built-in support for English, Mandarin Chinese, and 90+ languages via the open-source multilingual Whisper model. Language is auto-detected per recording.
WIDGETS
Add a VoiceSumm widget to your Home Screen or Lock Screen to see your most recent summary at a glance — also generated entirely on-device.
SHORTCUTS + SIRI
"Hey Siri, record a voice note with VoiceSumm" opens the app ready to capture. Or pick an audio file and run "Summarize with VoiceSumm" — Shortcuts integration built in.
REQUIREMENTS
• Requires iPhone 15 Pro / Pro Max or newer, or an iPad with an M-series chip.
• ~2.5 GB install size — most of which is the bundled AI models that make the offline summary possible.
OUR PROMISES
We will never:
- Add a free-with-ads tier
- Ship cloud sync or team features that require an account
- Sell, share, or collect your audio or transcripts
The promise is the product.
