LingTube Privacy Policy

1. Data We Process

Video, caption, and website content

• Caption or subtitle text, translated subtitles, generated SRT text, and SRT files you choose to upload.
• Video page metadata needed to provide the feature, such as video ID, video URL, video title, platform name, caption track metadata, current playback time, and nearby caption context.
• Temporary speech-recognition input derived from the current video only when you use no-caption caption generation. It is used as an intermediate processing input for caption creation and is not presented by LingTube as an exported, saved, or user-facing output.

Account and authentication data

• An anonymous device ID generated by the extension, session tokens stored locally, and encrypted device headers sent to LingTube services.
• Backend account records such as user ID, invite code, standard credits, AI credits or pass/subscription status, credit events, video usage records, and session records.
• If you link Google sign-in, LingTube verifies your Google access token with Google and stores your Google account ID, email address, email verification status, name, profile picture, linked device record, and related timestamps.
• If email/password authentication is used where available, LingTube stores the email address and a salted password hash, not the raw password.

Extension settings and local learning data

• Preferences such as UI language, source and target languages, translation service, TTS engine, selected voice, rate, pitch, volume, subtitle display mode, subtitle style and position, quick translation, and speech-recognition toggles.
• OpenAI-compatible provider settings, including provider, endpoint, model, scene, and API key. The API key is stored in your browser and is sent only to the provider endpoint you select when that feature is used.
• Local AI Learn data such as saved review cards, XP/streak statistics, due dates, lesson cache, and answer review state.
• Local counters and caches such as daily normal-start usage, voice catalogs, translation memory, and temporary audio cache tokens.

Technical and fraud-prevention data

• IP address, IP hash, approximate country from Cloudflare headers, user agent, extension version, request ID, error/status information, and timestamps.
• Payment or access-pass metadata when you purchase or activate paid AI features, such as order ID, product ID, payment status, amount, credits/pass type, creation time, and paid time. LingTube does not store full payment card numbers.

2. How We Use Data

• To read captions from supported video sites, translate them, display them, and read them aloud.
• To generate subtitles for videos without captions when you choose that workflow.
• To provide AI translation, AI Learn lessons, answer grading, and video chat features when you request them.
• To keep your settings, local learning progress, account session, credits, invite rewards, and paid access working across sessions.
• To prevent abuse, enforce device/IP/account limits, avoid duplicate credit charges for the same video, diagnose failures, and maintain service security.

LingTube does not sell user data, does not use user data for advertising, and does not use or transfer user data for creditworthiness or lending decisions.

3. Chrome Extension Permissions

LingTube requests the following Chrome extension permissions only for the product features described in this policy:

• activeTab: used to work with the current supported video tab after you open or start the extension. It helps LingTube locate the active video page and apply subtitle/voice controls for that tab.
• scripting: used to inject LingTube content scripts and helper scripts into supported video pages so the extension can read available caption tracks, place the subtitle overlay, coordinate playback, and run optional speech-recognition caption workflows.
• storage: used to store extension settings, local account/session state, local learning data, voice and lesson caches, daily usage counters, and temporary metadata in your browser.
• offscreen: used for background audio and TTS work, including Microsoft Edge TTS WebSocket synthesis and Azure Speech synthesis, without requiring a visible extension page.
• declarativeNetRequest: used only for Microsoft speech-service compatibility rules. The rules adjust request headers for Edge/Azure TTS endpoints so speech synthesis works reliably; they are not used to block, redirect, or inspect arbitrary browsing traffic.
• cookies: used only for the Microsoft Edge TTS service requirement described in the Cookies section. LingTube may read or set an anonymous muid cookie for speech.platform.bing.com and does not read unrelated website cookies.
• identity: used only when you choose Google sign-in. LingTube requests Google OAuth scopes openid, email, and profile to verify and link your account.

LingTube does not use these permissions to collect browsing history, monitor unrelated websites, inject ads, or track users across the web.

4. Host Permissions and Domains

LingTube requests host permissions for the domains below so the extension can provide caption reading, translation, speech synthesis, AI features, account access, and optional no-caption workflows. Access is limited to the feature purposes listed here.

Supported video platforms and caption/playback sources

• https://www.youtube.com/*, https://youtube.com/*, https://youtu.be/*, https://www.youtube-nocookie.com/*: read YouTube caption tracks, video metadata, watch-page data, and page context needed for supported caption workflows.
• https://www.coursera.org/*, https://coursera.org/*: read Coursera video captions and playback context.
• https://www.udemy.com/*, https://udemy.com/*, https://*.udemycdn.com/*: read Udemy lecture/caption metadata and caption files from Udemy CDN.
• https://www.bilibili.tv/*, https://bilibili.tv/*, https://www.bilibili.com/*, https://bilibili.com/*, https://api.bilibili.tv/*, https://data.bilibili.tv/*, https://api.bilibili.com/*, https://data.bilibili.com/*, https://*.bilivideo.com/*, https://*.hdslb.com/*, https://subtitle.bilibili.com/*, https://*.bstarstatic.com/*: read Bilibili subtitle metadata, subtitle files, and playback resources needed for supported caption workflows.
• https://www.tiktok.com/*, https://tiktok.com/*, https://*.tiktok.com/*, https://*.tiktokcdn.com/*, https://*.tiktokv.com/*: read TikTok caption metadata, WebVTT caption URLs, video metadata, and playback resources needed for supported caption workflows.

Translation, AI, and speech services

• https://translate.googleapis.com/*: send subtitle text for Google Translate when normal translation is selected.
• https://speech.platform.bing.com/*, wss://speech.platform.bing.com/*, https://*.tts.speech.microsoft.com/*: fetch voice lists and synthesize speech through Microsoft Edge TTS or Azure Speech. Text/SSML, language, voice, rate, and pitch may be sent.
• https://api.openai.com/*, https://api.groq.com/*, https://openrouter.ai/*, https://generativelanguage.googleapis.com/*, https://api.anthropic.com/*: send subtitle text, context prompts, provider settings, model name, and your locally stored API key only when you enable OpenAI-compatible translation with that provider.
• https://nmlg-server.johndyl456.workers.dev/*, https://lingtube.xyz/*: handle anonymous device bootstrap, sessions, Google account linking, credits, invite rewards, AI translation, AI Learn, video chat, SSML generation, short-lived Azure token issuance, payment access links, and abuse prevention.

Authentication and account verification

• https://accounts.google.com/*: opened through Chrome Identity only when you choose Google sign-in.
• https://oauth2.googleapis.com/*, https://www.googleapis.com/*: used by LingTube backend services to verify the Google access token and retrieve the Google profile fields needed for account linking.

Optional no-caption speech recognition

• https://*.supabase.co/*, https://api.uniscribe.co/*: create a temporary speech-recognition session, process intermediate recognition input through signed URLs, start transcription, poll task status, and export generated SRT when you use optional online caption generation.
• https://*.r2.cloudflarestorage.com/*: access signed cloud storage URLs or temporary files used by optional no-caption speech-recognition processing.
• Verified third-party media analysis microservices, when configured for no-caption caption generation: prepare temporary speech-recognition input for the current video. LingTube uses this only to create captions and does not expose the intermediate processing content as a user-accessible output.

5. Local Storage

The extension uses browser storage such as chrome.storage.local, page localStorage, IndexedDB, CacheStorage, and in-memory caches.

• Settings, local auth state, AI Learn cards, voice cache, lesson cache, and daily usage counters remain in your browser until you clear extension/browser data or uninstall the extension.
• Temporary recognition-input caches are used for no-caption caption generation and are deleted when the tab closes or navigates where possible.
• OpenAI-compatible API keys are stored locally in the browser and are not uploaded to LingTube's backend by the normal OpenAI-compatible translation flow.

6. Network Transfers and Third Parties

Data is sent only as needed to provide features you use. Depending on your settings and actions, data may be transferred to:

• Supported video sites, including YouTube, Coursera, Udemy, Bilibili, and TikTok, to read caption tracks, video metadata, or media needed for the requested feature. Site cookies may be included by the browser for same-site requests, but LingTube does not read or store those cookie values.
• Google Translate (translate.googleapis.com) for normal subtitle translation.
• Microsoft Edge TTS and Azure Speech endpoints (speech.platform.bing.com and *.tts.speech.microsoft.com) for voice lists and speech synthesis. Text or SSML, selected voice, rate, pitch, and language may be sent for synthesis.
• Your selected OpenAI-compatible provider, such as OpenAI, Groq, OpenRouter, Google Gemini-compatible endpoint, or Anthropic, when you enable OpenAI-compatible subtitle translation. Subtitle text, context prompts, model settings, and your API key may be sent to that provider.
• LingTube backend and AI services for account/session handling, credits, AI translation, AI Learn, answer grading, video chat, SSML generation, short-lived Azure token issuance, payment access links, invite handling, and abuse prevention.
• VividScribe/UniScribe, Supabase, and signed cloud processing URLs when you use optional online caption generation. Temporary recognition input, file metadata, language, task IDs, and generated SRT may be processed by those services.
• Verified third-party media analysis microservices, when configured for no-caption caption generation. They receive only what is needed to prepare temporary speech-recognition input for the current video, and LingTube does not present the intermediate processing content as a user-accessible output.
• Google OAuth endpoints when you choose Google sign-in.
• Payment processors when you open or complete a paid AI access flow. Payment processors handle payment details under their own policies; LingTube stores only the order/access metadata needed to provide the paid feature.

Third-party services process data under their own privacy policies and terms.

7. Cookies

LingTube uses the Chrome cookies permission only for the Microsoft Edge TTS service requirement. The extension may read or set an anonymous muid cookie for speech.platform.bing.com so Edge TTS requests work reliably. LingTube does not use the cookies permission to read unrelated website cookies.

8. Retention

• Local extension data remains until you clear browser data, reset the extension, or uninstall it.
• Server account, session, invite, credit, subscription/pass, payment metadata, and video usage records are retained while needed to operate the service, prevent abuse, provide support, and keep accounting records.
• Session records expire according to server configuration. Logout revokes the active session token.
• Caption, SRT, temporary recognition input, and AI request content is processed for the requested feature. Some AI translation results or request metadata may be cached to avoid repeated processing or duplicate credit charges.
• Third-party providers may retain data according to their own policies.

9. Security

LingTube uses HTTPS for network transfers where supported. Passwords are stored as salted hashes. Session tokens are stored as hashes on the backend. IP/device data may be hashed for abuse-prevention limits. Google access tokens are used to verify and link your account and are not stored as long-term LingTube account credentials.

10. Your Choices and Controls

• You can use browser built-in voice instead of cloud TTS where available.
• You can use Google Translate, your own OpenAI-compatible provider, or LingTube AI features depending on your settings and subscription/access state.
• You can avoid Google sign-in, avoid optional AI features, or avoid no-caption speech recognition if you do not want those data transfers.
• You can clear extension data in Chrome, reset browser site data, log out, or uninstall the extension.
• For account deletion or privacy questions, contact johndyl456@gmail.com.

11. Changes

We may update this policy when LingTube changes its features, data handling, or service providers. The effective date at the top shows when this policy was last updated.