Everything You Need to Know.

Press your hotkey, speak, and text appears at your cursor. Works in any app on your Mac — Slack, Docs, Mail, Cursor, ChatGPT, everything. That simple.

You can tap, hold, or double-tap the hotkey depending on your preference. Text is automatically pasted wherever your cursor is. No copying, no switching windows.

macOS 14 or later. Optimized for Apple Silicon (M1-M4) but works on Intel Macs with GPU acceleration. 4GB RAM minimum.

Larger Whisper models benefit from more RAM and faster chips, but the smaller models run great on any supported Mac.

Yes. 14-day free trial via TestFlight. Full app, no limitations. No auto-charges or commitment. Just download and start talking.

Because oto isn't the main event — your voice is. It's a quiet tool that shows up when you need it and disappears when you don't. No splash screens. No branding in your face. Just a vessel for your words.

The name itself means "sound" in Japanese. Lowercase felt right for something that's meant to stay out of the way.

100% local. No cloud. No telemetry. No account. We can't see your data even if we wanted to.

All three transcription engines run entirely on your Mac's hardware. Your audio never touches a server.

Never. All three engines (Whisper, Parakeet, Apple SpeechAnalyzer) run on your hardware. Zero network requests for transcription.

If you use optional AI enhancement with a cloud API key you provide, only the text is sent for cleanup — never your audio.

No. No analytics SDK. No Firebase. No Mixpanel. No telemetry of any kind.

Apple handles payments through the App Store and tells us basically nothing identifiable. We don't even know your name.

100+ languages with automatic detection. Speak in any supported language and oto transcribes it. No need to manually select a language — it figures it out.

Yes. Whisper models can translate any supported language to English. Speak Japanese, get English text. Perfect for ESL professionals who want to skip the mental translation.

Parakeet v3 and Apple SpeechAnalyzer output text in whichever language you speak — no translation.

WAV, MP3, M4A, AIFF, AAC, FLAC, CAF, MP4, and MOV. Drag and drop any file. Still private. Still offline.

Whisper models — Range from tiny to large. Larger models are slower but more accurate, and support more languages. The only engine that translates to English. 100+ languages.

Parakeet v3 — Fast with solid accuracy. Supports 25 languages. No translation — outputs text in whichever language you speak.

Apple SpeechAnalyzer — Apple's new on-device engine (requires macOS 26). Dramatically fast — benchmarks show 55% faster than Whisper. Growing language support with automatic updates.

All three run 100% locally on your Mac. No cloud. No internet required.

YES. Always. Transcribe on a plane, in a bunker, anywhere. Internet is only needed for the initial download from the App Store.

Because we think renting your voice is absurd. One month of the competition buys oto forever.

Voice-to-text runs on your hardware using open-source AI models. There's no server to maintain, no cloud to pay for. So why would we charge you monthly? We wouldn't.

Yes. No subscription. No upsells. No "Pro" tier. All features included. Free lifetime updates.

Current pricing is Founder's Pricing — the regular price will be $30. Lock it in now.

14-day refunds through Apple. Or try free via TestFlight first — full app, no limitations, no risk.

Yes. One purchase covers up to 5 family members through Apple's Family Sharing. Everyone gets the full app.