Press the mic and talk. Your words land in the box, after what you already typed, ready to send. It runs on your computer, so it is quick, private and works offline.
Try it: press the mic Β· a demo, nothing is recorded
The first time, Hedgehog asks to download its speech model, once. After that it starts at once.
The bars follow your voice and a clock runs. Pause, think, go on: it keeps the thread.
The words appear where the caret was. Edit them or send them as they are. Esc cancels.
No. Dictation runs on your computer with a speech model Hedgehog downloads once (about 570 MB). Your voice is turned into text locally, works offline, and only the text you send goes to the chat.
It detects the language by itself: English, Russian, and dozens of others. You can switch languages from one message to the next without changing any setting.
Press the mic in the message box and talk. The bars follow your voice. Press Enter to finish and the words land in the box after what you already typed, ready to edit or send. Esc cancels.
Yes, on Windows 10 and 11 as well as on Macs with Apple Silicon.