How it works
How a link gets checked
Six steps from pressing the button to the result. Pick a sample link and press Start.
For now the link is just text. It isn’t opened, so nothing can happen yet.
01
You paste the link
For now the link is just text. It isn’t opened, so nothing can happen yet.
02
Your browser takes the link apart
The main domain is picked out and scam signs are looked for: the @ trick, look-alike letters, piles of subdomains. Nothing is sent to a server at this step.
03
The link becomes an identifier
VirusTotal recognises links by a base64url identifier. First we check whether a finished report already exists for that ID.
04
The request goes through our server
Your browser talks to the CheckUrl server, and the server talks to VirusTotal. The API keys live only on the server and never reach your browser.
05
70+ engines answer
Kaspersky, ESET, Google Safe Browsing and other databases judge the link independently. For a brand-new link the analysis takes 30 seconds to 2 minutes.
06
The verdict is ready
The answers are counted: two or more “malicious” means dangerous; one means suspicious; none means no threats found. Every engine’s answer is shown in the report.
Good to know
How are files checked?
First, the file’s SHA-256 fingerprint is computed on your device. If VirusTotal knows it, you get the result instantly and the file itself is never sent.
Why does a new link wait?
A link VirusTotal hasn’t seen yet is checked by 70+ systems in turn. That usually takes 30 seconds to 2 minutes.
What does the AI do?
The AI doesn’t decide anything. It only explains the engines’ result in plain words, in one or two sentences.
Now try it yourself
Paste a suspicious link into the field on the home page — the same process runs with real data.