The noScribe software is very well designed for automatically generating transcription. It saves a great deal of typing and is truly a big help — thank you for providing such a useful tool!

In addition to the widely used AI speech recognition packages Whisper and Faster-Whisper (both by OpenAI), there are other free, open-source solutions that offer valuable possibilities. Two examples are:

VOSK: https://alphacephei.com/vosk/models

Parakeet: https://huggingface.co/nvidia/parakeet-tdt-0.6b-v3

It would be very helpful if these models could also be integrated into the respective software or apps in the future.

Furthermore, it would greatly enhance accessibility if speech-to-text models not only converted spoken language into written text, but also automatically noted background sounds — such as music, traffic noise, or knocking — in brief bracketed comments.

Thank you for your continued development work and dedication. It is truly appreciated!