What stays on your device
PageNarrator extracts webpage text, selected text, and PDF text inside your browser. Its built-in Kokoro voices also run inside the extension. Reading position, voice choices, and settings are stored in extension-local storage.
Voice files download when needed
Preparing a built-in voice or starting a read downloads missing model files from Hugging Face and its file-delivery hosts. They receive your IP address and the requested file paths, not narration text. After a successful download, the same voice and processing mode can reuse cached files offline. Switching voices or modes may need more files; clearing or evicting the cache requires another download.
Optional custom voices and online PDFs
Resume stores page URLs, titles, positions, and a short text excerpt locally. Imported model files run on your device. An endpoint voice sends narration text and voice settings to the local server or third-party API you configure, along with an optional API key for authentication. That provider's privacy policy applies. To keep speech on your device, use the built-in Kokoro voices. Online PDFs are requested from their document host. See the privacy policy for storage and deletion details.
What PageNarrator does not receive
PageNarrator servers do not receive the text you read, PDF contents, selected passages, page URLs, or browsing history for narration. The public website does not include first-party analytics or advertising scripts.
What the original website may handle
Your browser may fetch an online page or PDF from its original website as it normally would. That website has its own privacy terms.
When PageNarrator hears from you
No account is required. PageNarrator receives personal information only when you contact support, such as your email address and message.
For the formal policy, read the Privacy Policy. To get PageNarrator, see the Chrome extension page.