Browser-side processing (Mic Test and similar)

Mic Test, and certain other lightweight features, run on your device. They’re light enough that we don’t need to send anything to the cloud. For these features, your audio stays in the browser tab and is never uploaded. We’ll always tell you, per feature, whether audio leaves your device.

Studio-mode runs

Studio-mode runs (chapter QC, mastering, voice casting) send your audio to our cloud GPUs to process the job. We do not sell that audio, share it with third parties, or use it for any purpose outside SpeechSyn.

Sentence-level training data (opt-out)

When you flag, regenerate, or correct a sentence in Studio or Casting & Direction, we keep the original generation, your correction, and the surrounding sentence (stripped of anything that identifies you) to improve our models for audiobook creators. That’s the loop that lets the SpeechSyn model get better at long-form narration, character voicing, and the pronunciation of invented names. Opt out any time from Dashboard → Settings → Data & training; opted-out runs are processed and discarded after billing.

Cast voices and saved assets

Cast voices, IPA pronunciation pins, mastering presets, and character samples are stored in your account and used only to run jobs on your behalf. They are never used to train our shared models without your explicit, per-asset consent.

Account data

Email, hashed password, and credit balance are stored in Supabase (EU or US region, depending on deployment). We do not sell personal data.

Cookies

A session cookie from Supabase for authentication. No third-party analytics or ad cookies on authenticated pages.

Your rights

Export or delete your account at any time from Dashboard → Settings, or by emailing privacy@speechsyn.com.