Subprocessors

Last updated: April 23, 2026

A "subprocessor" is a third-party company that processes user data on behalf of fio to deliver the core service. We publish this list so users, reviewers, and data protection authorities can see exactly who touches your data and why. This disclosure supports our obligations under GDPR Article 28 and CCPA.

Each subprocessor below is under a Data Processing Agreement (DPA) or equivalent contractual terms that limit the use of your data to delivering the service described.

Current subprocessors

Name Location Purpose Data shared DPA
Supabase United States Hosting, authentication, database, file storage, edge functions Account data (name, email, auth tokens), saved URLs, generated episodes, transcripts, audio files, cover art DPA
Jina AI United States Extract clean text from URLs you save The URL you saved (no account identifiers) Terms
Anthropic United States Claude model for podcast script generation Extracted article text (no account identifiers). Zero-retention API; inputs not used for model training. Commercial Terms / DPA
Microsoft Azure United States / European Union Primary text-to-speech (Azure Neural Speech) Generated podcast script text (no account identifiers) DPA
OpenAI United States Fallback text-to-speech and fallback cover art generation Generated podcast script text (for TTS) or a short image prompt derived from the episode (no account identifiers). Zero-retention API; inputs not used for model training. DPA
Google United States Primary cover art generation (Gemini / Imagen) A short image prompt derived from the episode (no account identifiers) DPA
Sentry Rolling out in v1.0 United States Crash reporting and error monitoring Crash stack traces, device model, OS version, app version, anonymized user identifier DPA
PostHog Rolling out in v1.0 United States / European Union Product analytics and feature usage Pseudonymous usage events (screens viewed, actions taken), device model, OS version, app version DPA