Internet Speed for 4K IPTV: What You Actually Need

"How fast does my internet need to be?" is the most common pre-purchase question we get, and the honest answer has two parts. The first part is a simple number. The second part is that the number is not the interesting bit — how consistent your connection is over hours, and how it behaves at 9pm rather than at 11am, matters considerably more.

Per-stream requirements

Our channels are encoded at four adaptive rungs. Standard definition runs at 1.5 to 3 Mbps, 720p at 3 to 5 Mbps, 1080p at 6 to 9 Mbps, and 4K at 14 to 22 Mbps. The recommended connection speed for each is deliberately higher than the bitrate, because a stream that exactly matches your available throughput has no room to absorb the normal variation in a domestic line.

In practice: 5 Mbps for standard definition, 10 Mbps for 720p, 15 Mbps for 1080p and 25 Mbps for 4K. Those figures are per stream and they assume the stated speed is genuinely available at the time you are watching, which is the part most people get wrong.

On-demand content is more forgiving than live at the same bitrate, because the player can buffer minutes ahead rather than seconds. A 4K film will tolerate a line that a 4K live sports feed will not. Our full bitrate table is on the features page.

Why stability beats headline speed

This is the single most important idea in this guide. A speed test measures a short burst, usually to a well-connected server nearby, at the moment you press the button. Streaming needs consistent throughput over ninety minutes, along the specific route between you and the streaming server.

A line that bursts to 300 Mbps at 11am and drops below 20 Mbps at 9pm — because your ISP's local segment is contended and every household in the street is streaming — will buffer 4K during exactly the fixtures you care about. A steady 40 Mbps line will not. Upgrading the first line to a 500 Mbps tier changes nothing, because the constraint is contention on shared infrastructure rather than the speed you are provisioned for.

This is why we ask customers to speed-test at the moment the problem occurs rather than sending us a screenshot from the afternoon. The afternoon figure is almost never the relevant one.

Multiple streams at once

Requirements add up. Two 4K streams need about 50 Mbps, three about 75 Mbps, four about 100 Mbps. Then add roughly 20% for everything else in the household — phones syncing, a games console downloading in the background, someone on a video meeting, a security camera uploading.

A realistic planning figure for a four-person household running two 4K streams simultaneously is 70 to 80 Mbps sustained. If your line delivers that at 9pm, you are fine. If it delivers 200 Mbps at midday and 35 Mbps at 9pm, you are not, and the fix is not a faster plan.

A useful compromise: run the main television in 4K and secondary devices in 1080p. Nobody watching on a bedroom TV or a tablet can tell the difference, and it roughly halves your total requirement.

How to test properly

Test at the time the problem happens. If it buffers at 9pm on Saturday, test at 9pm on Saturday. A midday result is a different measurement of a different network condition.

Test three times and take the lowest. Single results vary substantially. The lowest of three is a better predictor of streaming behaviour than the best of three, which is what people instinctively quote.

Test on the device you actually watch on. A result from a laptop next to the router tells you nothing about the Fire TV Stick behind the television two rooms away. Most streaming platforms have a speed-test app; use it.

Test wired and wireless. If the wired result is 90 Mbps and the wireless result at the television is 18 Mbps, you have found your problem and it is not your ISP.

Watch for jitter, not just speed. Any test that reports jitter is worth attention: consistently under 10ms is good, over 30ms suggests a connection that will produce intermittent stalls even at a decent average speed.

Wi-Fi is usually the bottleneck

In our support data, local wireless problems account for more buffering reports than every other cause combined. The 2.4 GHz band has three non-overlapping channels and is shared with neighbouring routers, microwaves, baby monitors and Bluetooth devices. In dense housing it degrades severely in the evening — the exact window when people watch television.

The 5 GHz band has far more channels and much less interference, at the cost of shorter range and weaker wall penetration. For a television in the same or an adjacent room to the router, 5 GHz is almost always the right choice, and switching to it resolves a large share of complaints outright.

Ethernet removes the variable entirely. Most streaming sticks accept a USB-Ethernet adapter for around $15, and for a primary television it is the most cost-effective reliability upgrade available. If buffering vanishes on Ethernet, you have definitively proved the cause was wireless, and you can decide between running a cable permanently or adding a mesh node near the TV.

Router settings that actually matter

Split your 2.4 GHz and 5 GHz network names. Many routers present both bands under one name and decide for you, often badly — devices cling to 2.4 GHz long after 5 GHz would be better. Separate names let you pin the television to 5 GHz explicitly.

Choose a clear wireless channel. A free Wi-Fi analyser app will show which channels your neighbours are using. Moving to a quieter one takes two minutes and occasionally transforms evening performance.

Enable QoS if your router has it, carefully. Prioritising the streaming device helps when the household is competing for bandwidth. Badly configured QoS can throttle everything, so change one setting at a time and test.

Change your DNS resolver. Some ISP resolvers are slow or route streaming requests to a distant edge. Setting 1.1.1.1 or 8.8.8.8 at the router costs nothing and sometimes improves channel-change times noticeably.

Reboot the router monthly. Unplug for thirty seconds. Consumer routers accumulate stale state, and this is a genuinely effective piece of maintenance rather than folklore.

Data usage per hour

At 22 Mbps, 4K uses about 10 GB per hour. 1080p at 8 Mbps uses roughly 3.6 GB, 720p at 5 Mbps about 2.2 GB, and standard definition about 1.3 GB. Four hours of 4K in an evening is 40 GB, which on an unmetered home line is meaningless and on a 100 GB monthly mobile allowance is most of your month.

If you are on a capped connection, cap the quality in the player rather than rationing your viewing. Watching 1080p instead of 4K reduces consumption by nearly two thirds, and on a screen under 55 inches at normal viewing distance most people cannot reliably tell the difference.

If your line is genuinely not fast enough

Be honest with yourself about this, because it is the one problem no provider can engineer around. If your connection cannot sustain 15 Mbps during the evening, IPTV will be a frustrating experience regardless of who you buy it from, and cable or satellite is the better choice — neither depends on your broadband at all.

Before concluding that, though, run the tests above properly. A great many lines that appear too slow are actually fine on Ethernet, and the problem was a 2.4 GHz wireless link the whole time. Test wired before you spend anything.

If you clear the threshold, the remaining variable is the provider. Our free 24-hour trial is the cheapest way to find out whether your specific line and our specific network get along — run it in 4K, during your evening peak, on the device you actually watch on. And if it does not hold up, tell us during the session rather than afterwards, because we can move you to a different one of our 14 edge locations while you are watching.

UK peak hours are not US peak hours

“Test at 9pm” is incomplete if you do not say whose 9pm. A UK household’s worst wireless hour is often Saturday 15:00–17:30, when every terrace is on the same 2.4 GHz band during football. A US East Coast household’s worst hour is more often Sunday evening. Test the hour you actually watch, in your timezone, on the device that will sit behind the TV — not a lunchtime speed-test badge on a phone in the garden.

drextv.com serves both markets on one login. The megabit numbers in this article do not change by country. The clock you use for the test does. If you are an expat watching a UK kick-off at local midnight, that midnight is your peak, even if the street outside is quiet.