Text to speech
Text-to-speech belongs inside the media workflow.
VoxNotch treats generated speech as production work: voice choices, scripts, outputs, consent framing, and review history stay attached to the same workspace as the source recording and transcript.
Voice design workflow with voice cards, waveform previews, and narration output.
Model
Local TTS and hosted voice AI are different promises.
| Question | Local-first VoxNotch path | Hosted voice platform |
|---|---|---|
| Where generation runs | In the runner and storage boundary you configure. | In the provider service path. |
| What is metered | Local compute, storage, and operation. | Usually characters, minutes, seats, or credits. |
| What needs review | Consent, provenance, output quality, model license, and storage. | Provider policy, workspace settings, and usage terms. |
Text-to-speech questions
Is VoxNotch a hosted voice marketplace?
No. VoxNotch is a private media workspace with local-first voice workflows when the operator configures the relevant engines and models.
Does local TTS remove every cost?
No. It removes default vendor character metering for local generation, but hardware, model licenses, storage, review, and support still matter.