About this speed test
This tool uses M-Lab's open NDT7 protocol, the same measurement technology used by many independent speed-test tools. Your browser opens a direct WebSocket connection to a nearby M-Lab server (chosen automatically by M-Lab's locate service) and exchanges data for a short download phase and a short upload phase, each capped at 7 seconds. No part of the actual speed test passes through this website's own infrastructure — it is a direct, client-to-M-Lab measurement.
Why results vary
Speed test results depend on your device, Wi-Fi vs. wired connection, other active devices on your network, your ISP's routing, and the specific M-Lab server selected at the time. Running the test more than once, or at different times of day, gives a more complete picture. For your current public IP and ISP details, see the What Is My IP checker.
How to get a cleaner measurement
Close large downloads, pause cloud backups, move closer to the Wi-Fi router, or use a wired connection before testing. Run download and upload tests at different times of day to see whether congestion affects your connection. VPNs can reduce measured speed because traffic takes a longer or more crowded route.
What speed numbers mean
Download speed affects streaming, browsing, and receiving files. Upload speed affects video calls, backups, and sending large files. Ping reflects round-trip latency and matters for calls, games, remote desktops, and interactive tools. A single test is a snapshot, not a guarantee of your ISP plan speed.
What this tool checks
The test measures browser-level throughput to an M-Lab measurement server. It reports download speed, upload speed, minimum round-trip latency, selected server location, test duration, and the public value visible during the run. These measurements reflect the path between your browser and that server, which can differ from a router test, provider app, or another testing network.
How the measurement works
M-Lab's locate service selects a measurement server, then the NDT7 protocol runs short download and upload phases over WebSocket. The page updates the gauges while data transfers. Because the browser performs the test directly, CPU load, Wi-Fi signal quality, browser extensions, and active tabs can influence results. The tool is best used as a practical comparison, not a billing-grade measurement.
Common use cases
Use this page to compare Wi-Fi and wired performance, test after changing routers, check whether a VPN slows the route, compare peak and off-peak hours, or document performance before contacting an ISP. It is also useful after troubleshooting DNS, IPv6, or WebRTC settings because privacy tools can change routing and latency.
Limitations
This is a browser-based measurement, so it can be affected by CPU load, Wi-Fi interference, battery-saving modes, other tabs, router quality, and the selected test server. It measures a real path at one moment in time, not the maximum capacity of your line under every condition.
Troubleshooting slow results
If the result is much lower than expected, repeat the test near the router, close streaming apps, pause backups, and try a wired connection if available. Test another device on the same connection. If only one device is slow, the issue may be Wi-Fi, CPU load, browser extensions, or background software. If every device is slow, compare with your ISP plan and test at another time of day.
Related tools
Use What Is My IP to see which provider and route are visible during the test. Use IPv6 Test when some sites feel faster or slower than others. Use VPN Guide to understand why encrypted tunnels can reduce throughput or increase latency.
Frequently asked questions
How does this speed test work?
It uses M-Lab's open NDT7 protocol. Your browser connects directly over WebSocket to a nearby M-Lab measurement server and exchanges data for a few seconds to measure download speed, upload speed, and latency.
Does this test use this website's own servers?
No. The speed measurement runs directly between your browser and a public M-Lab server. This website does not relay, proxy, or store the transferred data.
Why did the test continue without location access?
Location access only helps pick a nearby server faster. If it is unavailable or denied, the test still runs using M-Lab's own network-based server selection.
Why is my result different from another speed test?
Different tests use different servers, routes, protocols, and durations. Browser load, Wi-Fi quality, VPNs, and congestion can also change the result.