SaaS
Utterbar
A private, offline dictation app for Windows. Fast on-device speech to text in over 30 languages, with nothing ever sent to the cloud.
Utterbar is a local dictation app for Windows. It turns your speech into text in real time and runs entirely on your own machine, so nothing you say is ever sent to the cloud.
The problem
Most dictation and speech-to-text tools depend on cloud services. That means your voice, which often carries personal, sensitive or work information, is streamed to someone else's servers. It also means latency, subscriptions and a hard dependency on a connection. I wanted dictation that was fast, private and fully under the user's control.
What I built
- A native Windows app that transcribes speech to text on-device, with no cloud processing.
- Support for 30+ languages, none of which ever leave the PC.
- A lightweight floating bar that stays out of the way and drops into any application.
- A small, self-contained download with a fast install-and-go experience.
- A clear privacy model where recordings and transcripts stay local by default.
Approach
The core principle is local first. Speech recognition runs on the user's own hardware, which keeps transcripts private, removes network latency and lets the tool work offline. The interface is deliberately minimal, a small bar you can drag anywhere, so dictation feels like part of the operating system rather than a separate app.
Why it matters
Utterbar sits at the intersection of privacy, productivity and desktop software. It shows how modern on-device speech models can replace cloud dictation without giving up speed or language coverage, while giving users full ownership of their data.
Current status
Utterbar is available for Windows 10 and 11 as a small, self-contained download, with ongoing work on languages, accuracy and workflow features.