Skip to main content

The Great Australian Speed Test

How does your internet rank?

Twenty five seconds. No signup.

Postcode only. Nothing else until after the test.

  • Sydney server
  • Measures bufferbloat
  • No email needed

Hold on, we're measuring

Warming up

0Mbps down

Right. Let's see what we're working with.

Here's your connection

—Mbps down
—Mbps up
—ms ping
—ms jitter
—Bufferbloat
—Your provider

We can't rank you yet

Tell us what you're on and we'll compare you against the right crowd.

What are you connected to?

What speed is your plan sold as?

Worth ruling out first: wifi distance, an old modem, and evening congestion. If it still reads the same tomorrow, it's worth a call.

Other plans worth a look

Tell someone

Your number means nothing to them. Your ranking they get instantly.

Your result card

OBBR Pulse

You just measured one moment. Want the whole day?

Pulse re-tests every 30 minutes in a popout window and charts your speed across the day, so you can catch the 8pm dip a single test always misses. No install. It needs an email, so your day of results is saved for you.

Your Pulse run is saved against this email so you can come back to the chart.

Track it over time

We'll remember this result and chart it against your next one.

For the most accurate result: turn off any VPN, close other tabs that are streaming or downloading, plug into ethernet if you can, and don't test on hotel or cafe wifi.

OBBR Pulse

One test is a snapshot. Pulse charts your whole day.

Re-tests every 30 minutes in a popout window, so you can see whether your connection drops at 8pm or all day long. No install.

Postcode so we know your area, email so your day of results is saved for you.

How this speed test works

Your browser opens a connection to our own server in Sydney, sends a controlled burst of data in each direction and times it. No third party widget, no server in Texas inflating your latency, no plugin to install. It takes about 25 seconds.

Then it does the part other speed tests don't: it works out who your provider is, asks two quick questions, and ranks your result against other Australians on the same connection type and the same plan. A gigabit line in a new estate and an FTTN line at the end of a long copper run are not the same product, and comparing them tells you nothing.

If the problem only shows up at night, OBBR Pulse is the tool for it. Start it from the box above, or from your results screen after a test. It re-tests every 30 minutes in a small popout window and charts your speed hour by hour, so you can see whether a slowdown is there all day or only at 8pm. No install, just an email so the chart is saved for you and you can send it to your provider.

Your postcode does two things. It puts your area on the postcode rankings once eight nearby tests exist, and as the data grows it lets us compare your result against your area rather than the national average. Providers advertise a maximum speed like NBN 100, but what you get when everyone in your street is streaming after dinner is usually lower. That evening figure is what matters day to day, not the number printed on the plan.

What this test measures

Download speed. How fast data reaches your house. The number on your plan, like NBN 100/20, is a maximum. This tells you what you actually get.

Upload speed. How fast data leaves your house. It matters more than people expect: video calls, photo backups and cloud sync all run on upload, and most NBN plans pair a much slower upload with the headline download.

Ping, or latency. The round trip time to our Sydney server. Lower is better. Under 20 ms feels instant, over 60 ms starts to hurt gaming and calls.

Jitter. How much that ping varies. A steady 30 ms is fine. A ping bouncing between 20 and 200 ms causes glitches even when the average looks acceptable.

Bufferbloat. The one that explains why a fast connection can feel slow. When the line is busy, some routers buffer too aggressively and ping balloons, so the speedometer reads 100 Mbps while your video call falls apart. We grade it A to F. Most Australian speed tests skip it entirely.

What counts as a good NBN speed in Australia?

Providers advertise a maximum, but the number that matters day to day is the typical evening speed, when everyone in your street is streaming at once. Here is what each NBN tier realistically delivers in the evening and what it comfortably handles.

NBN 12 10 to 12 Mbps

One person browsing and SD streaming

NBN 25 22 to 25 Mbps

A couple or small household, HD streaming

NBN 50 42 to 48 Mbps

Family of four, 4K streaming and video calls

NBN 100 85 to 95 Mbps

Heavy users, several 4K streams, gaming

NBN 250 220 to 240 Mbps

Large households, file servers, 8K streaming

NBN 1000 500 to 700 Mbps

Power users. Almost always overkill in 2026

If your result sits well under the evening figure for your tier, that is worth chasing. If it sits at or above it, you are getting what you pay for and a faster plan probably won't fix whatever is annoying you.

Why one test isn't the whole story

You test when something annoys you, which is rarely when the problem is worst. The NBN's hardest hours are roughly 7pm to 11pm, when the whole street is streaming. A test at 2pm can read beautifully and tell you nothing about the thing you actually noticed last night.

That's what OBBR Pulse is for. It re-tests every 30 minutes in a popout window and charts the whole day, so you can see whether the drop is constant or only at peak. The resulting chart is also the single most useful thing you can put in front of a provider's support team. "It feels slow sometimes" gets nowhere. "It drops to 22 Mbps every night between 8 and 10, here are twelve tests" is a fault report.

From Dan

I've been running this site for over 20 years and built this speed test in house so it could be properly joined up with the rest of OBBR. Your result goes into our database, gets attached to whichever provider we detect you're on, and shows you how you compare with other Australians on the same setup. The other reason for building it was bufferbloat. Most speed testers do download and upload fine and skip bufferbloat entirely, which is the thing that actually makes a fast connection feel sluggish in real use. We grade it A to F. If you find anything here that doesn't add up, let me know.

Frequently asked questions

What does this speed test measure?

Download, upload, ping, jitter and bufferbloat. That last one is the silent killer of fast connections and most tests skip it.

Why does it use a Sydney server?

Many speed tests bounce off servers overseas, which inflates latency and undersells your connection. Our target sits in Sydney, so what you see is your line rather than the Pacific cable.

What is bufferbloat?

What makes a fast connection feel slow when it's busy. Some routers buffer too aggressively under load and ping balloons, so the speedometer says 100 Mbps while your video call glitches.

How do you know which provider I'm on?

From your connection's network identity, checked across the ISP name, the organisation, the network operator and the autonomous system number together. Checking them together matters, because brands like Belong and Exetel run on other companies' networks and a single field check files those customers under the wrong company.

Can I trust the result?

It's a genuine measurement of your connection at that moment, from your device, over whatever wifi or cable sits between you and the router. That last part is the honest caveat: if you test on wifi three rooms away, you're partly measuring your wifi. Ethernet gives you the line itself.

How does this compare to Ookla or Fast.com?

The download and upload numbers should land in a similar range. The differences are that we measure bufferbloat, we test against a Sydney server, and we rank your result against other Australians on the same technology and plan rather than just handing you a number with no context.

Is it free, and do I need an account?

Free, no signup, no app. We ask for your postcode before the test so the comparison means something, and two optional questions after it. Nothing else.

What happens to my data?

Your result is stored so it can be counted in the rankings and so you can come back to it. Results from VPNs, datacentres and monitoring runs are excluded from every ranking. We publish the sample size next to every figure, and the full methodology explains what we exclude and why.

More on Australian broadband

Methodology: read how we measure and how we rank providers. Server location: Sydney. Test server status: live.

Self submitted results, not a statement about any provider's network. Plan tier and connection type are what you tell us. 140 comparable results in the last 90 days. How we measure.