Internet Speed Test
Powered by M-Lab NDT7. Free, open, unbiased.
Results are measured using the open-source M-Lab NDT7 protocol and are not influenced by any ISP.
Data rate from internet to your device. Affects browsing, streaming, and downloads. Most plans are defined by this number.
Data rate from your device to the internet. Critical for video calls, live streaming, and sending large files.
Round-trip time in milliseconds. Under 20 ms is excellent. Gamers and video callers care most about this number.
Variation in ping. Low jitter means your connection is consistent. High jitter causes choppy audio and frozen video.
What Is a Good Internet Speed?
Use this table to see how your results compare to typical household needs.
| Download Speed | Best For | Rating |
|---|---|---|
| Under 25 Mbps | Basic browsing, email, single-device streaming | Below Average |
| 25 – 100 Mbps | HD streaming, video calls, 2–3 simultaneous users | Average |
| 100 – 500 Mbps | 4K streaming, gaming, 4–6 simultaneous users | Good |
| 500 Mbps+ | Large households, home offices, smart home devices | Excellent |
Ping under 50 ms is good for most uses. Under 20 ms is excellent for gaming. Jitter under 10 ms is acceptable; under 5 ms is ideal for video calls. For a full breakdown, see our guide to what is a good internet speed.
How to Get an Accurate Result
A speed test only measures what it can see from your device to the M-Lab server. Small things can significantly affect the number you get.
- Use a wired connection. Plug an Ethernet cable directly from your router to your computer. Wi-Fi adds overhead and can easily cut measured speed in half.
- Close other applications. Background downloads, streaming, and cloud sync consume bandwidth during the test and will lower your result.
- Test at different times. Run the test in the morning and again in the evening. If evening results are significantly lower, your ISP is experiencing peak-hour congestion.
- Restart your router first. Routers accumulate stale connections over time. A quick reboot often improves results by 10–20% on older hardware.
- Run the test three times. Take the average. A single result can be anomalously high or low due to temporary server load or routing hiccups.
What Affects Speed Test Results?
If your result is lower than your plan speed, one or more of these factors is usually the cause.
Every wall, appliance, and neighboring network weakens your Wi-Fi signal. Testing over Ethernet eliminates this variable entirely.
Peak hours (7–11 pm) strain shared infrastructure. Speeds on DSL and cable can drop 30–50% during these windows.
A router more than 5 years old may not be capable of passing your full plan speed. This is the most common cause of disappointing results on fast plans.
DSL speed degrades with distance from the telephone exchange. Old or corroded copper wiring also causes signal loss that no upgrade can overcome without a line repair.
Why We Use the NDT7 Protocol
Most speed test tools route your traffic to a server inside your own ISP's network. This inflates results because the data never actually reaches the open internet. The number looks good, but it does not reflect what you experience when loading a website or joining a video call.
This speed test uses the NDT7 protocol, developed and maintained by Measurement Lab (M-Lab), a non-profit research consortium supported by Google, the Open Technology Institute, and Princeton University's PlanetLab. NDT7 routes your test to servers outside your ISP's infrastructure, giving a result that reflects real-world performance.
All test results are published as open data under M-Lab's research mission. No personally identifiable information is collected. Your IP address is anonymised before publication. See our privacy policy for full details.