Privacy policy

TranscribeVid turns video and audio into text, right in your browser. This page explains what that involves for your data, in plain words. In short: Your file never leaves your device. Your transcripts are kept in this browser only, there is no account, and the site sets no cookies of its own. Ads appear only on information pages and beside finished transcripts, and are served by Google, which uses cookies.

1. Who runs this website

TranscribeVid (transcribevid.com) is operated by Advaith Bibhishan Sonne, Chinnapanahalli Main Road, Bengaluru, Karnataka 560037, India ("we", "us"). We decide how and why your data is processed, as described here.

2. What is processed, and where

• Media you add. The video or audio files you choose or record. They are processed in your browser on your device (using WebGPU or WebAssembly). Your file is never uploaded to our servers and never sent to Google or anyone else.
• Text made on your device. The transcript, its timestamps and your edits, the title you choose, and the settings you pick (quality, language, vocabulary hint). They are made and kept in this browser; we never receive them.
• Requests that deliver the website. Cloudflare (Cloudflare Pages) delivers this website. As our hosting and network provider it processes your IP address and the usual request data (such as the time, the page requested and your browser) to deliver and protect the site.
• One running total. When a transcription finishes, your browser sends one anonymous "+1" to our counter (a Cloudflare Worker). It carries no file, transcript or identifier. Cloudflare sees your IP address in transit and uses it briefly to limit abuse (a rate limit); we store only the total. The number shown in the page header is reported by visitors' browsers and is not verified.
• A technical log in this tab. The Logs page shows this tab's own technical log. It stays in your browser and is never sent anywhere.

We do not ask for your name, e-mail address or any other account details.

3. Why it is processed

To give you the transcription you asked for on your own device, to deliver the website securely, and to keep the service free through advertising. Where data protection law such as the GDPR applies, our legal bases are providing the service you requested (Art. 6(1)(b)), our legitimate interest in security, preventing abuse and the anonymous running total (Art. 6(1)(f)), and, for advertising cookies where the law requires it, your consent (Art. 6(1)(a)), which Google's consent message asks for.

4. How long it is kept

• Your files and transcripts stay in this browser. Transcripts, edits and the source file are kept in this browser's storage (IndexedDB) only. They are deleted automatically when your session ends, which happens after 30 minutes of inactivity, or at once when you press Delete my data now in the Transcripts menu.
• Speech models stay in your browser's cache (Cache Storage), so the next transcription starts faster, until you delete them (Delete my data now can delete them too) or your browser clears its storage. They are program files, not your data.
• We keep nothing of yours on our side: apart from the single running total, nothing about your use is stored by us.

Because your transcripts exist only in this browser, export what you need before you leave. Deleted data cannot be recovered.

5. Downloads your browser makes

To transcribe on your device, your browser downloads the speech model files from Hugging Face (huggingface.co) and the speech engine's program files from jsDelivr (cdn.jsdelivr.net), and keeps them in its cache (Cache Storage) for next time until you delete them. Those providers receive the usual request data (your IP address, your browser and which file was requested), never your media or your transcript.

6. Advertising

TranscribeVid is free because it shows ads, served by Google AdSense. Ads appear only on information pages (home, How it works, FAQ and the tool pages) and beside a finished transcript. No ad is shown while a file is being transcribed, and never on the privacy, terms or logs pages.

• Third-party vendors, including Google, use cookies to serve ads based on your prior visits to this website or other websites. Google's use of advertising cookies enables it and its partners to serve ads to you based on your visit to this site and/or other sites on the Internet.
• You may opt out of personalised advertising by visiting Google Ads Settings, or opt out of some third-party vendors' use of cookies for personalised advertising at www.aboutads.info. Learn how Google uses information from sites that use its services.
• We never send your file, your transcript or any personal information to Google.
• In the EEA, the UK and Switzerland, and in some US states, Google's certified consent message asks for your choice when ads load.

Ads are not needed for any feature: with ads blocked or declined, everything works the same.

7. Cookies and browser storage

• No cookies of our own. The site itself sets no cookies.
• Google's cookies. Google sets its own cookies when ads load, on the pages that show ads only (see section 6).
• Storage in your browser. Your browser keeps your transcripts and files (IndexedDB) and the speech models (Cache Storage) as described above, remembers your light or dark theme choice, and its session storage keeps a few convenience settings for the current tab (such as your last quality choice). None of this is sent to us.

8. Sharing

We do not sell your data. We hold none of your files or transcripts, so we have none to share: the recognition runs on your device, not on an external transcription or AI service. Google (advertising), Cloudflare (hosting and the counter), Hugging Face and jsDelivr (the downloads) receive only what is described above. We may disclose information if the law requires it.

9. Security

Your files are processed in your browser and stay on your device. The website, the downloads and the counter use encrypted connections (HTTPS). Anyone who can use this browser profile can see the transcripts stored in it until they are deleted, so press Delete my data now on a shared computer.

10. Children

TranscribeVid is a general-audience service and is not directed to children under 13 (or under 16 in the EU where applicable).

11. Your rights and choices

You are in control: delete your transcripts, files and, if you like, the downloaded speech models at any time with Delete my data now. It deletes the data in this browser only; it cannot delete Google's cookies. To change your ad choices, use Google Ads Settings, or your browser's cookie settings. Because we hold no account and no details that identify you, we usually cannot link any data to a person. Where the law gives you rights over your personal data, such as access, correction, deletion, restriction, objection and portability, or the right to complain to your data protection authority, you can exercise them by contacting us: theadvaithsonne@gmail.com.

12. Open-source components

Some features download open-source components into your browser when you use them. They run on your device like the rest of the engine. Their licences ask us to show these notices:

• FFmpeg (ffmpeg.wasm core 0.12.10) (GPL-2.0-or-later). A 9 MB converter that runs in your browser. It is downloaded only when a file needs it (AVI, WMA, FLV, AMR and other formats your browser cannot open) and only after you agree. It is never part of the website itself. This converter is FFmpeg (https://ffmpeg.org) compiled to WebAssembly by the ffmpeg.wasm project, licensed under the GNU General Public License version 2 or later. It is used unmodified. Its complete corresponding source code, including the build scripts and the FFmpeg and library sources it is built from, is available at https://github.com/ffmpegwasm/ffmpeg.wasm/tree/v0.12.10; on request we will also provide a copy. Source code · Licence text
• Qwen3-ASR 0.6B / 1.7B (ONNX exports) (Apache-2.0). Speech-recognition models downloaded into your browser when you choose Qwen3 ASR (0.9 GB or 1.95 GB, kept in the browser cache until you delete them). Qwen3-ASR by the Qwen team, Alibaba Cloud, licensed under the Apache License 2.0. The files used are ONNX conversions of Qwen/Qwen3-ASR-0.6B and Qwen/Qwen3-ASR-1.7B by jiangzhuo9357 (encoder, embeddings, tokenizer) and sorryhyun (decoder), also Apache-2.0; they are used unmodified, pinned to fixed revisions and checked by hash. Source code · Licence text

13. Changes to this policy

If we change this policy, the new version appears on this page with a new "Last updated" date.

See also the Terms of use, How it works and the FAQ.