๐ŸŽฌ
CapMagic Studio
Privacy

How to Caption Videos Privately, Without Uploading Them Anywhere

Updated September 24, 2026 ยท 3 min read

Most online captioning services work by uploading your video to their servers. That's fine for a cat video, but not for unreleased client work, internal company recordings or anything under an NDA. Here's how browser-based captioning avoids uploads entirely, and what to expect from it.

How cloud captioning usually works

  1. You upload the full video to the provider's servers.
  2. Their servers transcribe it (and may keep the file for a while).
  3. You download captions or a captioned video.

That means waiting for large uploads, trusting the provider's storage and retention policies, and sometimes paying per minute.

How in-browser captioning works

Modern browsers can run AI models and process video themselves. CapMagic Studio uses that:

  • Speech recognition runs OpenAI's Whisper model inside your browser (using WebAssembly, in a background Web Worker), so your computer does the transcription.
  • Audio is decoded from your video file locally.
  • Export draws the captions onto each frame and encodes a new MP4 using your browser's built-in video encoder (WebCodecs), again on your device.

Your video file is read from your disk into your browser's memory, and the result is saved back to your disk. It's never sent to a server.

What is downloaded

Your browser still downloads the app itself and, the first time you create captions, the speech model files (roughly 45 MB). The model is stored in your browser's cache so later sessions start faster. None of these downloads contain your data; they flow in one direction only, from the website to you.

Technical users can confirm this with the browser's developer tools (Network tab): during captioning and export you'll see only downloads of app and model files, never an upload of your video.

Trade-offs to know about

  • Your device does the work. A recent laptop is much faster than an older phone; long videos take longer.
  • Limits: videos up to 250 MB and 10 minutes.
  • Language: automatic transcription currently supports English speech; other languages can be pasted or imported as text.
  • Browser support: MP4 export uses modern browser features (best in Chrome and Edge). Other browsers automatically use a fallback that may produce WebM.

Good fits for private captioning

  • Client work before it's approved or published.
  • Internal training, all-hands and meeting recordings.
  • Journalism, research interviews and legal material.
  • Anyone who simply prefers not to upload their personal videos.

Tips for the smoothest experience

  • Use a recent version of Chrome or Edge for the fastest MP4 export.
  • Close other heavy tabs and apps when processing long videos, so your browser has enough memory.
  • Keep the laptop plugged in for long jobs; some devices slow down on battery.
  • The first run is the slowest because the speech model is downloaded once. After that it loads from your browser's cache.
  • Keep your original video. Exports are new files; your original is never modified.

Frequently asked questions

Is my video stored anywhere?

No. The video is read into your browser's memory for processing and is gone when you close the tab. The only file created is the one you download.

Do you store my captions?

No. Captions exist only in your browser while you edit. Download the SRT file if you want to keep them.

Why does the page download files if nothing is uploaded?

Your browser needs the app and the speech model to do the work locally. Those are downloads from the website to you; your data never travels the other way.

Does it work offline?

You need an internet connection to open the website and, the first time, to download the speech model. The processing itself happens on your device.

Ready to caption your video?

Free, in your browser, and your video never leaves your device.

Open CapMagic Studio โ†’