SniffVid is a small, independent downloader that turns a video link into a file you actually own. It was built to prove that the heavy half of the job — merging audio and video — does not need a server at all.
When a download stops
Four guides cover the failures this site can explain: an unreadable source address, a bot check that never finishes, an expired link, and a browser that cannot write the file.
What happens when you paste a link
Your browser sends the address to a Cloudflare Worker. The Worker asks a handful of parsing services for the media streams behind that page and returns the formats they expose, each wrapped in an encrypted, short-lived token. The Worker never stores your link.
Where your media goes
Nowhere but your device. If the source CDN allows the browser to read it directly, SniffVid reads from the origin. If it does not, the media is streamed through the Worker as an opaque relay — it is never written to disk or to a queue, and the relayed bytes are discarded the moment your browser receives them.
Why merging happens locally
Merging is a byte-level copy of already-encoded tracks, so a browser can do it with the File System Access API and WebAssembly. Doing it locally keeps large files off our infrastructure, removes the cost of hosting them, and means a download cannot leak your viewing habits to us.
How the project is funded
SniffVid offers free downloads and guides. Pages may include clearly labeled tool promotions, with UTM parameters identifying the referral source. Visiting or installing a promoted tool is optional and is not required to use the downloader.
Who writes the tutorials, and how they are checked
The tutorials are written and reviewed by the people who build and run SniffVid rather than by an outside writing pool. Every article carries two dates — when it was published and when its steps were last run against the live parsers — so you can judge how current an answer is before you follow it.
- Each step is performed on a desktop browser and on a phone before it is written down.
- Where an article states a limit, that limit was reproduced rather than assumed.
- File sizes and bitrates are given as ranges, because every platform chooses its own.
- Claims that can be checked elsewhere are cited to vendor documentation or a specification, never to another downloader.
Corrections
When a step here no longer matches what you see, that is worth telling us. The contact page lists exactly what to include — the link you pasted, the message you saw, and your browser — and a correction that holds up is published as an edit to the article, with its review date moved forward.
- An article that names a platform is rewritten when that platform changes how it serves media.
- A limit that turns out to be a bug is removed once it is fixed, rather than quietly dropped.
- Promotions are clearly labeled, while tutorial capabilities and limitations are updated according to the available evidence.