Last updated · July 31, 2026
Privacy Policy
This policy describes what data CENDRO LABS PTY LTD collects when operating Scribix, how we use it, and the choices you have. Plain-English summary first; details below.
The Short Version
- We collect your email and name to create your account.
- We store the audio/video you upload and the transcript we make from it. Audio is deleted after 14 days; transcripts stay until you delete them.
- We process your audio only to generate and deliver transcripts for your account.
- Our browser extension sends the YouTube page URL you choose to process to Scribix so we can retrieve available captions. Paid AI summaries also send the caption text to Scribix for summary generation.
- When you use Ask AI, Scribix sends the transcript, your question, and recent conversation context to OpenAI. Scribix saves the conversation until you clear it, delete the transcript, or delete your account.
- We don't sell your data.
- You can delete any transcript at any time. You can delete your whole account from the account page.
What We Collect
- Account: email, name, avatar, and account identifier used for sign-in.
- Content: audio and video files you upload or record, YouTube URLs you ask Scribix to process, available YouTube caption text, and the transcripts and summaries produced from them. This also includes the questions and answers you save through Ask AI.
- Subscription: billing status, billing cycle, and current plan. Payments are handled by Paddle, and we never see or store your full card number.
- Usage: minutes consumed in the current billing period, file size, Ask AI questions consumed from your plan allowance or one-time free trial, browser-extension quota identifiers, IP-based abuse-prevention data, and basic logs needed to operate the service (timestamps and request status). For AI features, we also record the feature, model, request outcome, input, cached input, output, and reasoning token counts, and the applicable provider price used to estimate cost. These cost records do not contain transcript text, questions, or answers.
How We Use It
- To operate the service: store your files, run transcription, return results.
- To retrieve available YouTube captions, generate a summary, or answer questions about a transcript when you request those features.
- To enforce usage limits and process billing.
- To debug, monitor for abuse, and improve reliability.
- To contact you about your account when needed (e.g. failed payment).
Processors
We use service providers to run Scribix: Cloudflare for hosting, D1 database, and R2 file storage; AssemblyAI for transcription processing; Google and NextAuth for sign-in; Paddle for subscription checkout, billing, tax, invoices, and payment management; OpenAI for paid AI summary generation and transcript question answering; Google Analytics, Microsoft Clarity, and Plausible-compatible analytics served from actone.app for product analytics and reliability monitoring.
Browser Extension
The Scribix browser extension runs on desktop YouTube watch pages in supported browsers. It reads the current YouTube watch URL and your browser language preferences so Scribix can retrieve a suitable available caption track. When you explicitly request a paid AI summary, the extension sends the caption snippets to Scribix and OpenAI processes them to generate the summary. After you authorize the extension, it uses a short-lived access token and a rotating, revocable refresh token for account requests. It does not receive or store your password or Scribix website session cookie.
The extension stores a random quota identifier and up to 20 recently retrieved transcripts in browser-local extension storage. Cached transcripts expire after seven days. The refresh token expires after 30 days at the latest; signing out asks Scribix to revoke it and removes the local tokens. Access tokens expire after 15 minutes. This local extension data is also removed when you clear the extension's storage or uninstall the extension. The extension does not contain advertising trackers or execute remotely hosted code.
Retention
- Audio & video: deleted automatically 14 days after upload.
- Transcripts: kept until you delete them or your account.
- Ask AI conversations: kept until you clear the conversation, delete its transcript, or delete your account. Clearing a conversation does not restore used question allowance.
- AI cost records: token counts, model, feature, request outcome, and the applicable provider price are retained for aggregate cost analysis. When you delete a transcript, its identifier is removed from these records. When you delete your account, both the account and transcript identifiers are removed. The records do not contain your transcript text, questions, or answers.
- Browser-extension cache: up to 20 transcripts are cached locally for up to seven days. Generated summaries may be cached securely by Scribix to return repeat requests without processing the same caption text again.
- Account record: kept while your account is open. Soft-deleted immediately on account deletion; operational copies are purged on a regular cadence as part of standard operations.
Your Rights
You can access your data via the dashboard, export transcripts as TXT/SRT/VTT, and clear an Ask AI conversation, delete any transcript, or delete your entire account. If you live somewhere with stronger rights (GDPR, CCPA, etc.), email us at hello@scribix.io and we'll honor them.
Security
Files are private by default and only accessible via short-lived signed URLs. Transport is encrypted (HTTPS). We don't see or store your sign-in password.
Children
Scribix is not directed to children under 13 and we don't knowingly collect their data.
Changes
We may update this policy. Material changes will be reflected on this page with a new "Last updated" date.
Contact
Privacy questions or requests: hello@scribix.io.