Privacy Guides
Local vs Cloud Text-to-Speech: A Privacy Guide
FreeTextoSpeech Team Machine translation draft
This translated draft is available for review and is not indexed or included in hreflang until a native-language editor approves it.
Local and cloud speech can sound similar while handling a script very differently. Choose based on the sensitivity of the text and the capability of the device.
Local processing
The model runs in the browser after its initial download. Text stays on the device, but memory use and performance vary.
Cloud processing
Selected text is sent to a hosted service. It may run better on older devices, but you should review retention and provider terms.
Check every stage
Document extraction, editing, generation, storage and analytics can have different privacy boundaries. Confirm each one.