TranscriptionSuite adds WhisperX, NeMo, VibeVoice

This is the kind of open-source update that turns a neat local app into serious infrastructure for privacy-first speech workflows. Instead of betting on one model family, TranscriptionSuite is becoming a practical front end for comparing and running the best local ASR stacks on your own hardware.

• –WhisperX support matters because alignment and diarization quality are often what separate toy transcription demos from usable production workflows
• –Adding Parakeet, Canary, and VibeVoice gives users real model choice across speed, language coverage, and diarization behavior instead of locking them into one backend
• –The model manager and parallel processing mode push the project toward “daily driver” territory rather than hobby-tool status
• –Keeping everything local is still the core value prop: private transcription without shipping sensitive audio to a cloud API