Aqua Voice

Aqua — talk to your computer, get perfect text.

The slowest part of your computer is the keyboard. You think in full sentences; your fingers deliver them one key at a time. Aqua closes the gap. Hold a key, speak, and your words appear as clean, formatted text in whatever app you're already in — Gmail, Slack, Notion, Cursor, Claude Code, Google Docs, even the terminal. No browser extension, no per-app setup, no copy-paste. It works everywhere you type, in about half a second.

This matters more than ever. In the AI era, the bottleneck isn't the machine — it's you. Cursor and Claude write code in milliseconds while you type the prompt one finger at a time. Aqua removes that friction: speak your intent at the speed of thought and let your tools do the rest.

What sets Aqua apart is the model underneath. Avalon is our own speech-to-text model — trained not on audiobooks and news like most dictation engines, but on the way people actually talk to machines: code, prompts, product names, and technical jargon. It ranks as the #1 proprietary model on the OpenASR leaderboard, and the difference shows. When 9to5Mac ran Aqua against built-in Apple Dictation on the same speech, Apple made 17 errors. Aqua made one.

Aqua does more than transcribe. Deep Context reads what's on your screen to get names and terminology right and shifts tone to fit where you're writing. A custom dictionary and standing instructions learn your vocabulary and style once, and remember. On iPhone, Voice Edit Mode lets you revise just by talking: "remove the filler words," "translate to Japanese," "make it lowercase." It speaks 49 languages and syncs across Mac, Windows, and iPhone.

Built by a small team in San Francisco, backed by Y Combinator, SOC 2 Type II certified, and used by engineers, writers, and founders who've quietly stopped typing. Start free; Pro is $8/month.

We've typed for 150 years. It's time to speak.