Speech Recognition software for dictation

I am looking for a speech recognition software that runs well under manjaro. One idea would be to combine VOSK with nerd-dictation. Is there any experience and has anyone used it successfully?

The aim is to dictate speech directly into the word processor or Telegram.

1 Like

Seems pretty straight forward.

Theres a package in the AUR.


Hi, nerd dictation looks great, but I am unable to install…
Any help?

AUR takes hours and fail with errors at end
pip3 fail because virtual env
pipx fail too

Is there a simple way to install this?

What are those errors?

AUR packages also require prerequisites … do you have them ?

Solved by installing python-vosk-bin from AUR (instead of using pip3), then follow the instructions : GitHub - ideasman42/nerd-dictation: Simple, hackable offline speech to text - using the VOSK-API.