Broadband Speed Test

Find out whether your home broadband delivers the speed you pay for. Test with a cable, test more than once, and compare with the typical speed on your plan, not just the headline figure.

Download

Test progress

Ready to test

  1. Ping waiting
  2. Download waiting
  3. Upload waiting
  4. Packet loss waiting

Test server

Cloudflare network

Nearest location, chosen automatically

Elapsed

—

Data used

—

Packet-loss check

—

Download
— Mbps
Not tested yet
Upload
— Mbps
Not tested yet
Ping
— ms
Not tested yet
Jitter
— ms
Not tested yet
Packet loss
— %
Not tested yet

What is being tested?

Download and upload speed, ping, jitter and packet loss, measured in your browser with real test traffic to Cloudflare’s network.

How the test works →

How long does it take?

Under a minute on most connections. On a fast connection a test can use a few hundred MB of data.

Testing tips →

Your results stay with you

We don’t store your results on our servers. Recent tests are kept only in this browser.

Privacy policy →

Share your result

This link contains your numbers and the test time. Nothing else about you or your network is included, and nothing is stored on our servers.

What a broadband speed test shows (and what it doesn’t)

A browser-based speed test measures the real-time throughput and responsiveness between your computer and the nearest edge test server. It reveals what your device actually experiences at that exact moment: download speed, upload speed, latency (ping), jitter, and packet loss.

However, a single browser test cannot conclusively establish an ISP service breach or contract violation. Your test traffic travels through your home Wi-Fi or Ethernet cable, your router’s processor, your modem, and your neighborhood node before reaching the wider internet. A low result can stem from local network interference, outdated device hardware, or background downloads just as easily as an ISP line fault.

Advertised speed vs. typical speed: Compare against the right number

Most residential plans are marketed with headline “up to” speeds (such as “up to 500 Mbps”). In practice, that headline figure represents an engineering maximum under optimal conditions. For a realistic baseline, compare your test results against your plan’s typical speed.

In the United States, Federal Communications Commission (FCC) rules require major internet service providers to display standardized Broadband Consumer Labels at point of sale. These labels disclose the typical download speed, typical upload speed, and typical latency expected during peak evening hours (typically 7:00 PM to 11:00 PM). If your plan advertises 300 Mbps but the Broadband Label states typical peak download speed is 240 Mbps, a wired result of 250 Mbps means your connection is performing as expected.

In other regions, national regulators publish similar benchmarks. In the UK, for instance, Ofcom’s Broadband Speeds Code of Practice requires participating ISPs to give customers an estimated speed range and a minimum guaranteed download speed at sign-up.

Reading results by connection technology

Different broadband delivery technologies behave differently under real-world conditions. Understanding what is normal for your connection type prevents confusing physical limitations with a network fault:

TechnologyTypical performanceCommon bottlenecks to watch
Fiber (FTTH) Consistently close to plan speed; upload frequently equals download (symmetric). Latency is typically under 15 ms. Local device networking ports and home Wi-Fi are almost always the ceiling rather than the fiber line. See our fiber speed test guide.
Cable (DOCSIS 3.1) High download capacity, but upload is asymmetric and significantly lower (often 10–35 Mbps). Latency typically 15–35 ms. Shared neighborhood node congestion during evening peak hours; loaded latency (bufferbloat) spiking when uploads saturate.
DSL Modest download (10–100 Mbps) and low upload. Speed depends heavily on copper wire distance from the central office or street cabinet. Physical copper line attenuation, degraded telephone wiring in the home, or line noise. Speeds drop steadily the further you are from the exchange.
5G Home / Fixed Wireless Variable speeds (50–300+ Mbps download); latency typically 30–65 ms. Performance depends on cellular signal and tower capacity. Gateway placement relative to windows; cellular deprioritization during heavy mobile traffic hours; physical obstructions.
Satellite (LEO vs. GEO) Low-Earth orbit (Starlink) typically delivers 25–150 Mbps with 30–50 ms ping. Geostationary services (Viasat/HughesNet) exhibit 600+ ms ping due to 22,000-mile orbit. Weather interference (heavy rain fade); physical sky obstructions (trees, roofs); satellite handoffs.

How to run a fair broadband test

To eliminate home bottlenecks and gather reliable data, test under clean conditions:

  1. Connect directly via Ethernet. Plug a laptop or desktop directly into a router LAN port using a Cat 5e or Cat 6 Ethernet cable. Wi-Fi results reflect room-to-room wireless degradation, not what arrives at your modem.
  2. Verify port negotiation speed. Ensure your computer’s network adapter links at 1,000 Mbps (Gigabit) rather than 100 Mbps. A damaged cable or legacy port will cap results at roughly 94 Mbps.
  3. Pause background bandwidth. Temporarily disconnect other household devices, pause game updates (Steam, Xbox, PlayStation), close cloud backup software (OneDrive, Google Drive, iCloud), and close other open browser tabs.
  4. Test at multiple times across several days. Run tests in the morning, afternoon, and peak evening (7:00 PM to 10:00 PM) over 3 to 5 consecutive days.

What to do if your wired speeds remain consistently low

If your clean, wired test results consistently fall 20% or more below the typical speeds on your plan’s Broadband Label across multiple days:

  • Power cycle your modem and router: Unplug both for 60 seconds to clear internal memory, force frequency retuning, and renew DHCP leases.
  • Inspect splitters and cabling: On cable connections, bypass unnecessary coaxial splitters between the wall and modem. On fiber, check that the fiber jumper to the ONT is not sharply bent or crimped.
  • Test with a second wired device: Confirm that an older computer’s antivirus, VPN, or background processor load is not the limiting factor.
  • Contact your provider with your log: Present your wired test records. Our guide on how to fix slow internet walks through the exact diagnostic checklist ISP technicians use so you can demonstrate you have already performed them.

Frequently asked questions

Why is my speed lower than my plan?

Plans usually list a maximum or “up to” speed. WiFi, older devices, network congestion at busy times and other people using your connection all reduce what one test shows. Test with a cable first to rule out WiFi.

What speed should I expect from my plan?

In the U.S., providers must publish a broadband label for each plan that lists the typical download and upload speed and latency. Compare your wired results with the typical figures, not just the headline number.

What should I send my provider if speeds are low?

A list of wired test results with dates and times, ideally across several days including evenings. Mention that you tested with a cable and with other devices disconnected. That rules out most of the usual first-line suggestions.

Sources

  1. Broadband Consumer Labels Federal Communications Commission

    U.S. providers must show typical speeds and latency for each plan.

    Checked October 2026

  2. 2026 Section 706 Report (FCC 26-55) Federal Communications Commission

    Keeps the U.S. fixed broadband benchmark at 100 Mbps download / 20 Mbps upload.

    Checked October 2026

  3. Ofcom (UK communications regulator) Ofcom

    Publishes UK home broadband performance research.

    Checked October 2026