Privacy Policy

Talk2Viet AI · Chrome MV3 Extension · Last updated 2026-05-31

1. What we collect

Sign-in identity — when you sign in with Google, we receive your email address and display name from Firebase Authentication. We use this to identify your account and personalise translations.
Session token — a random opaque string stored in an HttpOnly cookie on talk2viet.vercel.app and mirrored in chrome.storage.local. It identifies your session to our backend.
Translation cache inputs — when you ask the extension to dub a YouTube video, we transmit the video's subtitle text plus its video ID to our backend so we can translate it and cache the result for everyone else. Subtitle text is YouTube's public data, not yours.

2. What we don't collect

Your YouTube watch history.
Any browsing data outside of youtube.com pages where you actively use the extension.
Your location, device pixels, microphone, camera, or clipboard.
Page content other than the subtitle track you ask us to translate.

3. How we use the data

Email + name — to display "logged in as …" in the popup and to look up your account on subsequent requests.
Session token — to authenticate your API calls.
Subtitle text + video ID — sent to an LLM (OpenRouter / Google Gemini) to produce a Vietnamese translation, and cached in our Postgres database so the next user who plays the same video gets an instant result. No per-user information is attached to cache rows.

4. Sharing

We share data with these processors strictly to operate the service. We do not sell or rent data to third parties.

Processor	Purpose
Firebase Authentication (Google)	Verifies your sign-in identity.
Neon Postgres	Stores user rows and the translation cache.
Upstash Redis	Short-lived session and stream metadata.
OpenRouter / Google Gemini	Generates translations from subtitle text.
Microsoft Edge TTS	Synthesises Vietnamese audio. Audio bytes flow directly from Microsoft to your browser — we never store them.

5. Permissions the extension requests

Permission	Why it's needed
`storage`	Keep your session locally so you stay signed in across browser restarts.
`scripting`	Inject the dubbing UI into YouTube watch pages.
`webRequest`	Capture YouTube's `pot` token from the network so we can fetch the subtitle track on your behalf. We never modify or block requests.
`notifications`	Display a one-time "audio is ready" toast.
`identity`	Open the Google sign-in window.
`host_permissions`	Required for youtube.com, talk2viet.vercel.app, accounts.google.com, identitytoolkit.googleapis.com, securetoken.googleapis.com to support the flows above.

6. Data retention & deletion

Your account row stays until you ask us to delete it. The translation cache row holds public subtitle text indefinitely (it contains no personal information). Email us at plusqnhaa@gmail.com to delete your account and all associated data.

7. Children

Talk2Viet AI is not intended for users under 13. We do not knowingly collect data from children under 13. If you believe a child has provided us data, please contact us and we will delete it promptly.

8. Changes to this policy

If the policy changes, the new version will appear at this URL with an updated "Last updated" date. Material changes will be announced inside the extension popup before they take effect.

9. Contact

Questions or data-deletion requests: plusqnhaa@gmail.com