Effective date: 2026-08-02
TranscriptLens (“we”, “us”) provides a Chrome extension that helps you search captions on YouTube videos you can play, ask optional grounded questions about spoken content, jump to timestamps, and copy sourced excerpts.
This policy describes what we collect, why, and how long we keep it. It applies to the TranscriptLens Chrome extension and the TranscriptLens API that the extension calls.
1. Summary
- Local-first. Transcript reading, keyword search, playback sync, copy formatting, and daily AI quota display are designed to run in your browser whenever possible.
- User-triggered AI. We process a video’s transcript and your question with an AI provider only after you choose Ask video and accept the first-use notice.
- Private by default. Anonymous product analytics are off until you turn them on. Analytics never include your questions, transcript text, copied content, or video identity.
- No public transcript library. We do not build a public index of video transcripts.
- No ads based on your viewing. We do not sell your personal data or use your viewing history for advertising.
2. Who this is for
TranscriptLens is intended for adults and professional users. The Signal MVP focuses on English captions on standard YouTube watch pages.
3. Data we process
3.1 Processed only in your browser (not sent for AI or analytics by default)
- Captions / transcript text for the current video (for reading, search, and copy)
- Your search query text (local keyword search)
- UI preferences such as privacy toggles and draft feedback text before send
- Local daily AI quota counters
3.2 Sent to TranscriptLens servers when you use optional features
| Feature | What we receive | Purpose |
|---|---|---|
| Ask video | Your question, the current video’s transcript segments needed to answer, and technical metadata (e.g. segment counts) | Generate a grounded answer and enforce daily limits |
| Installation / quota | Opaque installation credentials and daily completed-AI counts | Abuse prevention and free daily AI limit (50 completed answers per browser per local day) |
| Anonymous analytics (opt-in) | Event names and allowed properties only (e.g. caption source bucket, duration bucket, answer type labels)—never questions, transcript text, copied text, titles, channels, URLs, or video IDs | Understand product reliability and activation |
| Optional profile | Role and content-type preferences you submit | Segment product feedback; not inferred from videos |
| Feedback | Message body; optional email; optional current video URL if you enable it; limited diagnostics fields | Fix bugs and prioritize improvements |
3.3 AI subprocessors
When you Ask, TranscriptLens servers send the necessary question and transcript content to DeepSeek (DeepSeek Open Platform) to generate a structured answer. DeepSeek’s processing may occur in jurisdictions outside your own, including China, subject to DeepSeek’s terms and policies.
We do not send Ask content to other model providers. We do not intentionally log Ask question or transcript bodies in ordinary application logs.
3.4 What we do not collect by default
- Full browsing history
- Video IDs, titles, channels, or URLs in analytics events
- Question text or transcript text in analytics events
- Payment information (Signal MVP has no payments)
4. How we use data
- Provide Search, Ask, evidence playback, copy, feedback, and optional profile features
- Enforce the daily AI completion limit
- Improve reliability when you allow anonymous analytics
- Respond when you leave optional contact email on feedback
- Secure the service (rate limits, abuse prevention)
5. Legal bases (where applicable)
Where GDPR or similar laws apply, we rely on: contract / requested service (features you invoke), consent (analytics; AI processing notice before first Ask; optional email/URL on feedback), and legitimate interests limited to security and abuse prevention, balanced against your rights.
6. Retention
Approximate retention for server-side records:
| Data | Retention |
|---|---|
| Anonymous analytics events | 90 days |
| Profile (from last activity) | 12 months |
| Feedback body | 12 months |
| Feedback email (after closed/shipped) | 90 days |
| Feedback video URL (after closed/shipped) | 30 days |
| Aggregates after deletion | Counts that cannot identify an individual |
Ask transcript/question payloads are processed to produce an answer and are not kept as a user-facing history or public library. Browser storage you control (preferences, drafts, local quota) remains until you clear site/extension data or uninstall.
7. Sharing
We share data only with:
- DeepSeek, as described above, when you Ask
- Infrastructure hosts that run
api.transcriptlens.app(currently Fly.io) under our configuration - You, when you copy content to your clipboard or export text yourself
We do not sell personal data. We may disclose information if required by law or to protect the security of the service.
8. Your choices
- Use Search and reading without Ask
- Cancel before accepting the first AI notice
- Keep anonymous analytics off, or turn them off later in Privacy
- Reset anonymous ID and request deletion from Privacy (server-side deletion follows our deletion credentials flow; some aggregates may remain non-identifying)
- Submit feedback without email or video link
- Uninstall the extension and clear extension storage
Contact: support@transcriptlens.app
9. Children
TranscriptLens is not directed at children under 13 (or the minimum age in your jurisdiction). Do not use the product if you are under that age.
10. International transfers
If you use Ask or other server features, data may be processed in the United States (API hosting) and by DeepSeek in other regions, including China. Where required, we use appropriate safeguards and disclosures; this draft should be reviewed before a public commercial launch.
11. Changes
We may update this policy. The effective date at the top will change. Material changes affecting AI processing or analytics will also be reflected in the product notices where practical.
12. Contact
Email: support@transcriptlens.app
Product: TranscriptLens — Video Transcript Search
API host: https://api.transcriptlens.app