Speech to text
Private speech-to-text for recordings you cannot casually upload.
VoxNotch turns audio and video into searchable transcripts on infrastructure you operate. The local STT path is built for sensitive interviews, session records, archives, captions, and automation where custody matters as much as text quality.
Local speech-to-text workflow with microphone, transcript, and storage boundary.
Evaluation
What private STT needs to show.
Timing and speakers
Transcription should keep enough timing and speaker structure to support captions, review, notes, exports, and later search.
Custody before convenience
The recording should not need to become hosted meeting data before it becomes useful text.
Workflow after the transcript
The transcript should feed notes, collections, captions, exports, reports, and automation instead of ending as a lonely text file.
Speech-to-text questions
Is speech-to-text local by default?
The default STT path is local faster-whisper running in your own stack. Optional cloud paths, if configured, should be treated as explicit egress.
Does VoxNotch charge per transcription minute?
The intended commercial model is a perpetual code license, not per-minute vendor billing for the local path. You still pay for the hardware, storage, and operation you choose.