ISPreview reported on October 26 that Nufibre has added a 5G home broadband package to its range. Plans start at £25 a month with a Wi-Fi 7 router, or £22.50 a month with a Wi-Fi 6 router, on a 24-month term and the VodafoneThree mobile network. The listed expected speeds are 100 to 800 Mbps.
A product like this puts two wireless links between a laptop and the internet. One runs from the cell tower to the router. The other runs from the router to the device. When a video call stutters, which one is to blame?
What was announced
Figures here are as reported by ISPreview. The router is a ZTE G5 Air (Wi-Fi 7) or a ZTE G5B2 (Wi-Fi 6), and delivery is next day with no engineer visit or new fixed line, subject to availability. Data is advertised as unlimited, with a 750GB monthly fair usage policy. ISPreview notes that mobile performance can vary.
That last point matters more than the headline range. A range of 100 to 800 Mbps tells you the best case and the worst case are far apart. It does not tell you how often a given home sits at either end.
Two wireless links, two kinds of variation
The cellular link is shared with other users of the same cell, so its capacity and delay can change through the day. The in-home Wi-Fi link is shared with other devices and affected by walls, distance and neighboring networks. Either can produce a slow or uneven connection, and both produce the same complaint: "the internet is bad."
A Wi-Fi 7 router raises the ceiling on the in-home link. It does not change what comes into the house. If the 5G side delivers an uneven feed, a faster router distributes an uneven feed faster.
Why consistency matters more than the range
Consistency of Service is the minimum speed divided by the maximum speed over a test. Two connections can share an average and differ hugely on this measure, and real-time applications such as calls suffer on low-consistency links.
Here is an illustrative scenario, not a measurement from any real home or product. A tester runs a sustained transfer in the same room, once over Ethernet to the router and once over Wi-Fi.
Wi-Fi: minimum 90 Mbps, maximum 480 Mbps, so 90 ÷ 480 = 0.19
In this example the wired result shows the 5G side delivering a reasonably steady feed. The Wi-Fi result, from the same line at the same time, swings widely. The problem sits inside the home, and no change of ISP would fix it. If the wired test is just as uneven as the Wi-Fi one, the problem is upstream, on the cellular link.
Round-trip time tells a similar story. Throughput on a TCP session is bounded by RTT, so a link whose delay swings will see its throughput swing with it. Our article on the importance of latency covers the mechanism in more detail.
What to measure
Before judging a 5G home connection, collect the same results under different conditions:
- Sustained throughput, RTT and consistency, not a single peak reading.
- Results at different times of day, since a shared cell can change with load.
- The connection type for each result, so Wi-Fi and wired tests can be separated.
- Jitter and packet loss for voice and video, which a speed figure does not show.
MyConnection Server sends real test traffic and measures sustained TCP throughput, RTT and consistency of service, and jitter, loss and MOS for simulated voice and video. Its connection tests detect whether the tester is on Wi-Fi or wired Ethernet and store that with each result. Currently on Windows, they also record Wi-Fi signal strength and quality. That is a quick indicator, not a Wi-Fi assessment, but it shows at a glance whether Wi-Fi was in play and how strong the signal was, so a poor result can be compared against the connection type. The same approach applies to remote workers on home connections of any kind.
For the full picture, see the whitepaper on why consistency of service matters.
Sources: product, price, contract term, router, network, speed and fair usage figures are as reported by ISPreview, October 2026. The 410/520 and 90/480 Mbps figures are an illustrative scenario, not measured data. Accurate as of November 2, 2026.