Internet Speed Test

Sends and receives test data to measure your connection — never your files, and nothing is stored

How it works

1. Close Everything Else That Uses the Network. Pause downloads, cloud backups, and streaming on this and other devices — they all compete for the same bandwidth the test is trying to measure.
2. Get Onto Ethernet, or Sit Close to the Router. Wi-Fi is the most common bottleneck in this whole test, not your internet line — a cable removes that variable entirely.
3. Press Start and Leave the Tab in the Foreground. Browsers slow down background tabs, which would make the result read artificially low.
4. Watch Latency First, Then Download, Then Upload. The test runs the three phases in order and updates the live number as each one runs.
5. Read the Result Against Your Plan, Not Against Zero. A number below your headline speed is normal — see below for exactly why.
6. Run It Again, at a Different Time of Day. One run measures one moment. A second run, especially in the evening, tells you whether the first was typical.

How to Get a Number You Can Trust

A speed test measures whatever is happening on your network at that exact second, so the conditions around the test matter as much as the test itself. A few minutes of setup avoids the most common reason for a confusing result.

• Nothing else downloading, uploading, or streaming on this device or others sharing the connection.
• Wired to the router if at all possible — Wi-Fi adds a variable this test can't separate out.
• The browser tab stays in the foreground for the whole run.
• No VPN active, unless you specifically want to measure the VPN's effect.
• Not running right after the router or modem just rebooted — give it a minute to settle.

Why a Browser Speed Test Reads Lower Than Your Plan

A gap between the number on this page and the number on your bill is normal, and it's rarely your internet line that's the limit. Four things usually explain it.

• Wi-Fi, not the line: the radio link between your device and the router has its own ceiling, and on a fast enough plan it's the actual bottleneck, not your ISP.
• CPU-bound encryption: every byte this test moves is encrypted, and decrypting it costs CPU time — an older laptop or phone can top out well below a gigabit line before the network is the limit at all.
• One connection vs. several: a single data stream over a real-world path can't fill a very fast line on its own; multi-stream tests (this one included) get closer, but a single download in a browser tab often won't.
• Distance and routing: you're measuring the path to a specific server, not your line in isolation — a closer or better-peered server reads faster than a distant one, even on an identical connection.

There's also a fixed, unavoidable gap: a line advertised as "1 Gbps" carries roughly 940 Mbps of actual data once Ethernet, IP, and TCP framing take their share. That alone accounts for a small shortfall on any connection, on any test, from any provider.

Mbps, MB/s, Latency, and Jitter: What Each Number Means

The four numbers this test reports each answer a different question.

• Download (Mbps): how fast data arrives at your device — the number that matters for streaming, browsing, and downloads.
• Upload (Mbps): how fast data leaves your device — the one that matters for video calls, cloud backup, and sending large files.
• Latency / ping (ms): the round-trip time to the test server. Lower is better, and it matters most for anything real-time.
• Jitter (ms): how much that round-trip time varies from one moment to the next. A connection can have low latency and still feel choppy on a call if jitter is high.

Wi-Fi Is Usually the Bottleneck, Not the Line

If your download number looks low but your plan is fast, the radio link is the most likely place to look before anything else.

• Band: the 2.4 GHz band is slower and more congested than 5 GHz or 6 GHz, but reaches further through walls — most routers use whichever seems reasonable by default.
• Channel width and congestion: neighboring Wi-Fi networks and channel overlap eat into the practical maximum, especially in apartment buildings.
• Distance and obstacles: a wall, floor, or 10 extra meters can cut throughput well before the connection drops entirely.
• The client's own radio: an older phone or laptop's Wi-Fi chipset can cap out far below what the router itself supports.
• Mesh backhaul: on a mesh system without a dedicated wireless or wired backhaul link, each hop between nodes can roughly halve throughput.

The fastest way to confirm it's Wi-Fi: run this test once wired and once wirelessly, from the same spot, a minute apart. A large gap between the two points squarely at the wireless link, not your internet plan.

What Speed Do You Actually Need?

Headline speed matters less than whether it comfortably covers what you actually do, at the same time, on the same connection.

• Web browsing and email: comfortable well under 5 Mbps.
• HD video streaming: roughly 5 Mbps per simultaneous stream.
• 4K video streaming: roughly 25 Mbps per simultaneous stream.
• Video calls: 3–4 Mbps down and up is comfortable; calls are far more sensitive to latency and jitter than to raw speed.
• Cloud backup and large uploads: bounded by your upload speed, which is usually much lower than download on most home connections.
• Online gaming: modest bandwidth, but low, stable latency and jitter matter more than the Mbps number for how the game actually feels.

That's why this test surfaces latency and jitter as prominently as the headline Mbps figures: for calls and gaming specifically, they're the better predictor of how the connection will actually feel.

When the Result Really Does Point at a Problem

Most low readings have an explanation above that isn't a fault. A smaller set of patterns is genuinely worth raising with your provider.

• Consistently under half your contracted speed, wired, across multiple runs and times of day — not just once.
• Jitter regularly above 30 ms on a wired connection, which usually feels like stuttering calls or games even when the download number looks fine.
• The connection reads fine at 3am but collapses every evening — that pattern points at local network congestion, worth raising with your ISP directly.
• A noticeably lower upload than download is normal by design on most home connections, not a fault on its own.

Frequently asked questions