WebRTC Leak Test
See which IP addresses WebRTC exposes in your browser, next to the IP websites see, with a plain verdict.
Running the WebRTC test (takes a few seconds)
Runs in your browser. The addresses aren't sent to us.
What the WebRTC leak test checks
WebRTC lets browsers make direct audio, video and data connections. To find a route, the browser collects its own network addresses and asks a STUN server which public address its traffic comes from. Any website can start that process with a few lines of JavaScript, without asking permission, and read the addresses.
This test does the same in your browser: it opens a throwaway data connection, collects the addresses, and compares the public ones with the IP our server sees. The STUN request goes from your browser straight to Cloudflare's public STUN server (stun.cloudflare.com), and the addresses WebRTC finds are never sent to us. Local addresses your browser hides behind a random .local name are shown as hidden.
Reading the result
No leak means WebRTC showed no public address, or only the one websites already see. A leak means it showed a different public address of the same type: if you use a proxy or VPN, that's most likely your own connection's address, and any site that checks can read it. An extra IPv6 or IPv4 address is flagged on its own: without a proxy that's normal on a connection with both, and with one it means that traffic isn't going through it.
Proxies set in the browser or the system usually carry web traffic only. WebRTC sends its own UDP traffic, which goes around an HTTP proxy, so a setup that looks right for websites can still leak here.
How to stop WebRTC leaks
Firefox: open about:config and set media.peerconnection.enabled to false to turn WebRTC off, or set media.peerconnection.ice.default_address_only to true to limit it to your default route.
Chrome, Edge and other Chromium browsers: there's no switch in the settings. Google's WebRTC Network Limiter extension, set to use your proxy server, or the WebRtcIPHandling policy on a managed Chrome, stops WebRTC from using routes that go around the proxy.
Brave: open Settings, then Privacy and security, and set WebRTC IP handling policy to Disable non-proxied UDP.
Safari: it hides local addresses from sites that don't have camera or microphone access, but has no setting for the public address. If Safari leaks, send all of the device's traffic through your VPN or proxy app rather than just the browser's.
Automation tools and multi-profile browsers usually have a WebRTC setting for each profile: set it to follow the profile's proxy or turn it off. After any change, reload this page and run the test again.
Questions
Is a WebRTC leak dangerous?
It shows the sites you visit the IP address they'd see without your proxy or VPN. It doesn't give anyone access to your device. Whether that matters depends on why you use the proxy.
Does the test need camera or microphone access?
No. It only opens a data connection, which browsers allow without asking. That's also why any site can run the same check.
Why does the test show a .local address?
Modern browsers replace local addresses with a random name ending in .local, so sites can't read your local IP. That's your browser protecting you, not a leak.
Will turning off WebRTC break anything?
Video calls and screen sharing that run in the browser need WebRTC. Limiting it to your proxy's route usually keeps calls working; turning it off stops them.
Related guides
- Web Scraping Anonymity: What Proxies Hide and Don't
Web scraping anonymity explained: what a proxy hides (IP, network, location), what still leaks (headers, DNS, WebRTC, fingerprints) and a 5-step check.
- Proxy vs VPN: Differences and When to Use Each
Proxy vs VPN: a VPN encrypts all device traffic, a proxy routes one app's traffic through another IP. Compare scope, speed and use cases.
- HTTP vs SOCKS5 Proxies: Which One Do You Need?
HTTP proxies handle only web traffic; SOCKS5 handles any TCP or UDP connection. Here is how to pick the right one for scraping, bots, or torrenting.