Privacy Policy

Last updated: September 7, 2026

The short version

We keep as little as possible. Your audio and video files are deleted right after processing. We don't sell data, we don't run ads, and we don't share anything with third parties for marketing.

What you send us

• Your media files (audio/video you upload or link). They are used only to generate your transcript and written assets. The upload is deleted automatically once processing finishes. A copy stays attached to your review page so you can play it while correcting — and it is removed when we rotate storage.
• Your email, city, and how you describe your speech (e.g. "English with a Cantonese accent"). These tune the recognizer to how you actually talk.
• Your corrections — every fix you make on the review page. These build your personal dictionary so the same mistake doesn't come back.

What we do with it

• Transcription runs through the Groq API (Whisper). Written assets run through Zhipu GLM. Files are sent for processing and are not stored by these providers beyond their standard processing needs.
• Your corrections feed your own dictionary only. If you join a regional word pool (3+ creators in the same city/language), shared words are contributed anonymously — city and language tag only, never your email or identity.

What we never do

• Sell or rent your data.
• Use your videos to train models.
• Share your email with anyone.

Deleting your data

Want everything gone? Email Paul.j.sutton@proton.me and we'll wipe your account, dictionary, and any stored media. No questions asked.

Cookies

We don't use tracking cookies or analytics. The site works without them.

Contact

Questions about any of this: Paul.j.sutton@proton.me