We built the tool
we couldn't find.
Tellaflow started because every voice dictation tool we tried either required an internet connection, charged a subscription, or did something uncomfortable with our audio. So we built one that doesn't.
It started on a plane. No WiFi. A 10-hour flight. And a voice tool that had exactly one thing to say: "Internet connection required."
We'd been using voice dictation on and off for years. The technology was always impressive. The products were always slightly off. They wanted to hear your calls, process your audio, store your transcriptions, bill you monthly. That's not a problem with voice recognition that's a business model problem.
Whisper changed things. When OpenAI released it, the obvious question was: why does voice recognition need a server at all? Your Mac has an M-chip with a Neural Engine that was literally designed for this. The latency of local inference turned out to be better than the round-trip to a server.
Tellaflow is the result of asking 'what if we built it right?' on-device, offline-first, free because we can be, privacy-first because the architecture demands it. We don't have grand ambitions about replacing keyboards forever. We just wanted to dictate a commit message at 30,000 feet.
What we believe
Ready to stop typing?
Free, forever. No account. No API key. Just your Mac and your voice.
Download Tellaflow free