Version 3.85 and later added a real-time speech-to-text function. To avoid excessive software size, the required models were not included. You need to manually download and extract the files, then place the extracted files into the models/onnx folder in the software directory.
Download from github.com
https://github.com/jianchang512/stt/releases/download/0.0/realtimestt-models.7z