TGRadioBooks
Book narration and text privacy: what is sent for processing?
A book uploaded to TGRadio stays in your personal library and is not automatically published. However, sections of its text are sent to a speech synthesis service for narration. Generation does not run entirely on your device. If text privacy matters, assess whether that transfer is acceptable before adding the book.
Review the source document first
Before uploading, open the book on your device and review its appendices, captions and inserted notes as well as the main text. Use material you have the right to use. Personal access to an uploaded book does not by itself establish whether sending its text for processing is acceptable.
For example, a prepared document might include personal comments containing contact details. If they are unnecessary for listening, remove them in an external editor before import. TGRadio has no built-in editor for text, pronunciation or SSML.
Separate personal access from text processing
Personal uploads do not automatically enter the shared catalogue. That describes whether other readers can access the book. It does not mean the text remains exclusively in your phone's memory. Narration uses a speech synthesis service outside your device.
Do not infer model-training terms, precise retention periods or the absence of other processing from the word “personal.” Those guarantees are not established here. If your document requires specific conditions, ask support about them before transferring its text.
Know when audio preparation begins
Adding a book extracts its text. You then choose the book and voice and press Play to begin preparing audio. Listening can start with the first ready section, while later sections are prepared as you listen. Importing does not automatically narrate the entire book in the background.
Text sections are sent for speech synthesis. Changing the voice requires a new audio version, so it is more than adjusting an existing recording. Playback speed is a separate setting and does not require new narration.
Distinguish your profile from device storage
Your listening position is shared through your profile. To continue on another device, use the same Telegram account. Browser sign-in does not require giving TGRadio your Telegram password: you confirm the sign-in through the bot when the codes match. Check the displayed code before confirming.
Offline audio is stored locally on the device where you saved it. A shared listening position does not transfer those downloads between devices. If you plan to listen on another phone, check the availability of the required sections on that phone.
Offline listening does not reverse text processing
Previously saved sections can play without a connection, but preparing them already required speech synthesis. Offline listening does not make generation local. New sections and new imports still require internet access, even when part of a book is already saved.
Signing out or clearing browser storage removes local audio. It does not confirm deletion of every copy held by the synthesis service or other systems. If you need specific deletion and retention arrangements, ask support concrete questions about the text, prepared audio and applicable retention periods before uploading the document.