dictatray

Speak into any window. Dictatray is a Windows tray application that turns your microphone into a keyboard: it listens, decodes, and types the words into whichever window has focus. Decoding runs on your own machine by default; nine cloud engines are available on your own API key.

npm install -g dictatray
dictatray

Before you install

• Windows only. The install refuses anything else.
• The binary is downloaded, not bundled. Installing fetches dictatray.exe (~71 MB) from https://dictatray.com and verifies it against the SHA-256 published beside it. A mismatch aborts the install.
• The binary is signed by KARDITORWEB, UNIPESSOAL, LDA — the company behind Dictatray. Right click dictatray.exe → Properties → Digital Signatures shows it. A certificate still earns its reputation with SmartScreen download by download, so an early version may still be asked about once.
• This is a beta. There is no stable release yet: every version published so far is a beta one, and that is what an ordinary install gives you.

What it does with your data

Local decoding never leaves the machine. If you configure a cloud engine, audio goes to that provider under your own key and their terms — Dictatray adds no service of its own in between, and holds no account for you.

More, including how it works and what it costs: https://dictatray.com