The Best Wi-Fi Mesh-Networking Kits
Photo: Michael Murtaugh
  1. Electronics
  2. Networking

The Best Wi-Fi Mesh-Networking Kits

  • We’ve retested the Eero 6 and Eero Pro 6 mesh kits after a major firmware update. Both mesh kits now perform satisfactorily, but we reaffirm our current picks.

There may be spots in your home where the Wi-Fi sucks, which is often more frustrating than no Wi-Fi at all. If you have a larger space, a mesh-networking kit may be the solution. Instead of using a single router, mesh-networking kits use multiple access points spread around your house to improve the range and performance of your Wi-Fi. After spending more than 200 hours evaluating and testing 24 mesh Wi-Fi networking kits in home and lab environments, we’re confident the Asus ZenWiFi AC (CT8) set is the best mesh-router choice for most people who need one.

Most people, however, don’t need mesh Wi-Fi and can do just fine in smallish homes and apartments with a regular router.

While many recent reviews tout Wi-Fi 6 as the solution to all your wireless problems, our tests and research led us to continue to recommend a Wi-Fi 5 (aka 802.11ac) mesh kit as our pick. Wi-Fi 5 (and some Wi-Fi 4) laptops, phones, and other devices will continue to dominate your home for the next few years, and the premium that Wi-Fi 6 demands just isn’t worth it yet.

Our pick

Asus ZenWiFi AC (CT8)

Asus ZenWiFi AC (CT8)

Best Wi-Fi mesh-networking kit

The ZenWiFi AC (CT8) is consistently fast in a variety of tests and surpasses more expensive options. It’s also easier to set up than more complex systems.

Buying Options

*At the time of publishing, the price was $300.

The Asus ZenWiFi AC (CT8) is a great choice for folks who want to set up a lag-free Wi-Fi network all over their home, as well as anyone who needs a mesh network that can better wrangle the growing number of smart devices in a home. It’s a bit slower overall than our upgrade pick, but that’s like saying an eagle isn’t as fast as a falcon, and it’s easy to set up and manage.

Advertisement

Upgrade pick

Asus ZenWiFi AX (XT8)

Asus ZenWiFi AX (XT8)

Ahead of the curve, for gigabit internet

The ZenWiFi AX (XT8) is almost a clone of our CT8 pick, but with significant upgrades in the form of future-proof technologies like Wi-Fi 6 and WPA3. We’d only suggest it if you have a gigabit (1,000 Mbps, or close to it) internet service plan.

Buying Options

$450 from Amazon

May be out of stock

*At the time of publishing, the price was $450.

The Asus ZenWiFi AX (XT8) is the right choice if you currently use or are planning on upgrading to a gigabit or faster internet connection. The XT8 is almost a mirror twin of our pick, the CT8, but it includes extras such as Wi-Fi 6 compatibility and WPA3 security. Neither are critical inclusions right now, but both will help keep the XT8 from becoming obsolete for a longer period of time. For nearly $100 more than the CT8, the speed improvements won’t be worth it for most people in most homes. But the upgrade is worth the added expense if you’ve already invested in gigabit internet service and need a powerful mesh network to work with all that bandwidth. It’s the upgrade pick for those who need the latest technology to squeeze out the fastest connection, always.

Everything we recommend

Our pick

Asus ZenWiFi AC (CT8)

Asus ZenWiFi AC (CT8)

Best Wi-Fi mesh-networking kit

The ZenWiFi AC (CT8) is consistently fast in a variety of tests and surpasses more expensive options. It’s also easier to set up than more complex systems.

Buying Options

*At the time of publishing, the price was $300.

Upgrade pick

Asus ZenWiFi AX (XT8)

Asus ZenWiFi AX (XT8)

Ahead of the curve, for gigabit internet

The ZenWiFi AX (XT8) is almost a clone of our CT8 pick, but with significant upgrades in the form of future-proof technologies like Wi-Fi 6 and WPA3. We’d only suggest it if you have a gigabit (1,000 Mbps, or close to it) internet service plan.

Buying Options

$450 from Amazon

May be out of stock

*At the time of publishing, the price was $450.

Before joining Wirecutter in 2018, Joel Santo Domingo tested and wrote about PCs, networking, and personal tech for PCMag.com, Lifewire, HotHardware, and PC Magazine for more than 17 years. Prior to that, Joel was an IT tech and system administrator for small, medium-size, and large companies.

Metaphorically, Joel has been a wire cutter for decades: Testing wireless home networking has been a part of his life for the past 20-plus years through all versions of Wi-Fi, back to the wireless phone extension he tacked onto the back of his Apple PowerBook. He did that so he could connect to the internet from his desk, his couch, and his bed seamlessly (a rarity for the time).

If you have a house that a single powerful router can’t cover well (probably some homes bigger than 2,300 square feet or so, depending on the layout), a large apartment or small house with signal-killing interior walls (such as lath-and-plaster, brick, stucco, or concrete blocks), or one with a very long or tall, narrow plan, like a three-story townhouse, you should consider a mesh-networking kit. But before you toss everything out and get a mesh kit, you should try moving your router to a central location—in smaller houses a single router can actually be more effective than mesh networking.

If you already have a good router that you like, and you need just a little more range in part of your house, you might consider adding a wireless extender. (Here's our comprehensive guide to wireless extenders.)

Another option are mesh extenders, which like mesh-networking kits, automatically hand you off from router to extender and back, using the same network name; that makes the mesh experience a little more seamless. The takeaway is that mesh extenders may improve coverage in dead spots if you already have a decent wireless router, though they showed mixed results in our extender guide testing compared with full-blown mesh networking kits.

A wired network is always faster than a wireless one

A graph showing the wired vs the Wi-Fi throughput.
Tested for an earlier version of this guide, a wired Ethernet connection was more than seven times faster than the highest throughput of any wireless network. No matter what misleading hype you may see on the box, Wi-Fi cannot outrun Gigabit Ethernet, period. Graph: Wirecutter
A graph showing the latency of wired vs Wi-Fi
The picture doesn’t get any prettier when you look at latency. The message here should be clear: If you really care about maximum performance, you need a wired connection, no matter how good your Wi-Fi gear is. Graph: Wirecutter

If your house is wired for Ethernet, you don’t need a mesh-networking kit. You can run Ethernet cable to inexpensive wireless access points and outperform the best mesh-network kits we cover here, at a much lower cost. Mesh shines when you don’t have wires, don’t want wires, and have lots of trouble spots (or one really big trouble spot) with poor or no coverage.

A mesh kit won’t necessarily make your internet faster at short to medium range. As shown in the performance testing of standalone routers, the best Wi-Fi mesh kits did just as well as our upgrade standalone router pick, the Asus RT-AX88U. Mesh can offer better coverage and lower latency in a wider area, which makes your connection feel faster throughout the house because your devices aren’t grabbing at faint wisps of signal.

A good standalone Wi-Fi router can handle multiple devices, as long as those devices all have good connections. Even one device with a poor connection can bring the quality of a single router’s network down, eating up all of the available airtime, starving the rest. The best mesh networks ensure good connections between devices, the base unit, and any satellites, reducing the situations where a poorly connected device can slow down the others. This reduces latency—the time you’re sitting waiting for the website, game server, or streaming service to respond and send data back. Replacing poor connections with better ones is what helps mesh shine.

But for spaces that a single router can’t cover, the most important thing about mesh-networking kits is how they’re designed to work together—the vendor has specifically tuned the units to establish fast, reliable “backhaul” connections with one another. That’s something you can’t get just by adding another access point or wireless extender to your existing router.

The Wi-Fi mesh networking kits we tested lined up side by side.
Photo: Michael Murtaugh

For past versions of this guide, we tested absolutely every mesh kit we could find—but these days the number of mesh kits is growing, and we’re getting pickier. We researched dual- and tri-band mesh kits from each of the major router manufacturers, including Asus, D-Link, Eero, Netgear, and TP-Link. We also looked for routers from lesser-known manufacturers with strong reviews from tech experts or potentially interesting features that set them apart.

For this round, we dropped some of the poorest-performing kits from our previous rounds of testing (such as the D-Link COVR-C1203, Linksys Velop dual-band kit, and Zyxel Multy U kits). We also dismissed kits that cost more than $500: If your home is particularly large or problematic, we’d recommend you look into hiring an IT professional to design and install a combination wired and wireless mesh network.

That left us with almost a dozen kits from eight vendors (Asus, D-Link, Eero, Motorola, Netgear, TP-Link, TrendNet, and Ubiquiti) to test, in order to find the best Wi-Fi mesh network kit.

We determined our recommendation using the 10 criteria below, drawn from research and extensive testing. To simulate the real-world activity of a busy home network, instead of testing maximum throughput from the mesh system to a single laptop, we used six laptops, spaced around 3,000 square feet of a three-and-a-half-story suburban home. We tested for speedy throughput (streaming simulated 4K video and file downloads), good coverage in spots around the house, and short latency (simulating three simultaneous browsing sessions on a busy network). We repeated each test set six times and averaged the results to smooth out spikes. See the How we tested section below for more details on our testing method and results.

We used similar criteria to our router reviews, tailored to fit our ideal mesh-networking kits:

  • Ease of setup and administration: Setting up a new network is often hard, but it doesn’t need to be. You should be able to get your home on the internet in less than half an hour with a mesh network.
  • Good speed test results: Speed claims on the box don’t mean much in the real world. In our tests, network speed, or throughput, varies from “This YouTube video will never finish loading” to “You can download a video game in an instant.” Generally, we looked for the mesh kits that performed above average, and we dismissed the slower routers.
  • Good range test results: You should be able to connect to a well-placed mesh kit from anywhere in your house. We tested each kit to see its maximum potential when close to the base unit, as well as in trouble spots in the home, to see how well the mesh signals could be successfully routed around obstacles like appliances and walls.
  • Low latency test results: Slow internet sucks. Latency—or lag—is the time spent waiting for the next thing to happen. A great mesh kit minimizes that wait even if the network is busy serving other devices.
  • Multiple Ethernet ports: Ethernet ports on a mesh kit’s satellites let you connect devices such as TVs, streaming boxes, and gaming consoles away from the base unit, giving them the benefit of unfettered access to the internet bandwidth you’re paying for. Ethernet ports on the satellites or nodes also let you extend the network with wires, using wired backhaul.
  • Expandability: You should be able to add more nodes later to extend and improve coverage even farther, if you discover dead spots or if you move to a larger house.
  • A fast processor and RAM: A router with a speedy multi-core processor and extra RAM can handle more connected devices and provide improved performance. The slow processors found in most cheap routers can still drag things down, no matter how good the radios that broadcast the Wi-Fi are. Not all manufacturers disclose the CPU and memory inside of their routers, but when we find outstanding specs, we do consider them when deciding what to test.
  • Nice-to-have extras: Fast, reliable Wi-Fi is what matters the most in a mesh kit, but more expensive optional features bring other benefits, too. The things we like to see that justify spending more for a mesh kit include speedier connections (like 2.5-gigabit ports and 802.11AX/Wi-Fi 6), built-in security (like WPA3), extra Ethernet and USB ports, VPN connections, and parental filtering.
  • Price: You can buy a mesh kit for $80; you also spend almost $1,000. But we don’t consider the cheapest or the fastest to be the best. When considering both features and our test results, we looked for “the best for the most for the least.” Right now, paying around $300 buys you excellent performance and features that offer real benefits. Spending 50% ($150) more buys small improvements for few people, and spending only half as much would mean, for most people, giving up a lot.

In addition, we consulted customer reviews on Amazon and Newegg, plus professional mesh kit reviews and performance rankings from CNET, Dong Knows Tech, PCMag, PCWorld, SmallNetBuilder, and Trusted Reviews, to determine our list of contenders. After identifying every model that met our criteria, we tested throughput, latency, features, and general user experience in a large test home.

It can be hard to buy a great router without testing them side by side because the connection speed classes stamped on the box don’t actually mean much. These classes come in the form of initially impenetrable alphanumeric jumbles such as “AC3200” or “AC1750.” In this case, “1750” stands for 1.75 gigabits per second. They refer to theoretical maximum ceilings defined in the specifications of wireless protocols, but have little to do with the speeds you’ll get in the real world.

One more thing: Don’t confuse the test results in our guide with the internet speed you’re paying for. For example, the Asus ZenWiFi AX is capable of throughput over 600 megabits per second at close range with no obstructions, but you still can’t get more than about 100 megabits per second from the internet if you have a 100-megabit plan from your ISP.

Our pick for the best Wi-Fi mesh networking kit the Asus ZenWiFi AC (CT8).
Photo: Michael Murtaugh

Our pick

Asus ZenWiFi AC (CT8)

Asus ZenWiFi AC (CT8)

Best Wi-Fi mesh-networking kit

The ZenWiFi AC (CT8) is consistently fast in a variety of tests and surpasses more expensive options. It’s also easier to set up than more complex systems.

Buying Options

*At the time of publishing, the price was $300.

The Asus ZenWiFi AC (CT8) is a great choice if you want to set up a lag-free Wi-Fi network all over a larger home, as well as folks who need a mesh network that can handle a growing number of smart devices in a home. In a variety of tests spread out around a three-story home, the CT8 was able to outperform all but our upgrade pick, including some kits costing considerably more, and it was relatively painless to set up.

If you’ve read multiple mesh reviews, our choice of a Wi-Fi 5 (aka 802.11ac) mesh kit as our pick might come as a surprise when many manufacturers and reviewers seem to be pushing beleaguered homeowners toward Wi-Fi 6. But we tested each of the mesh kits with a mix of Wi-Fi 6 and Wi-Fi 5 devices—just like you’re likely to have in your home for a few years yet—and the Wi-Fi 6 kits we tested didn’t match the CT8’s good speed and excellent coverage.

The raison d'être of a mesh network is to provide good Wi-Fi performance all over the house, not just near the base unit. The CT8 was one of the fastest mesh kits in our test scenario. It was just behind its more expensive sibling, the ZenWiFi AX (XT8), at close range and in the attic where the test laptop was all but guaranteed to connect through a mesh point. And it actually topped the XT8 in one of the bedrooms that proved a particularly tough testing spot.

A graph comparing the stacked throughputs of multiple mesh networking kits.
We tested the mesh kits ability to push data as fast as they can at three spots in the house: one nearby, with no obstructions; a spot where they were almost guaranteed to use the mesh node through one jump; and a spot that requires the signal to go through multiple walls. Graph: Wirecutter

The laptop in the bedroom in question (WC3 in our graphs) was challenging because it was close enough to connect to a slower 2.4 GHz signal through multiple walls and appliances between it and the base unit—but WC3 could grab a faster 5 GHz signal if it chose to connect through a satellite or mesh node. This isn’t an unusual use case in homes that would particularly benefit from a mesh setup, but only 4 of the 11 mesh kits tested well in this situation, with the CT8 on top. Two notable kits, the Netgear Orbi RBK752 and Ubiquiti Dream Machine with BeaconHD came close, but both cost at least $100 more than the CT8.

A bar graph comparing the median latencies of each mesh networking kit.
Each bar represents the time a computer was waiting for tasks to finish in our tests. Longer bars represent longer waits, in which you and others on your network would be staring at a load screen. Graph: Wirecutter

In our overall latency tests, the CT8 finished right behind its pricier sibling, the ZenWiFi AX CT8. In these tests, raw speed in the form of throughput, in megabits per second (Mbps), isn’t the only thing that matters. Our latency test showed how routers handled the added stress of multiple devices accessing the network simultaneously. High latency, or lag, can make an otherwise speedy connection seem to drag, especially when the network is busy. The four finalists bunched at the top of this test were all tri-band kits that use their third radio band only to communicate with other network nodes, not your actual phones or laptops. That factor alone wasn’t a guarantee of success, however: The Netgear Orbi RBK752 was a tri-band kit that performed more like two of the dual-band kits (the Deco X20 and the Eero Pro with 2 Beacons).

The Asus ZenWiFi AC (CT8) next to its AC adapter.
The Asus CT8’s minimalist styling extends to its AC adapter. Photo: Michael Murtaugh

A great router or mesh kit like the CT8 is relatively easy to set up, even if you don’t know much about technology. We try to avoid kits that have inscrutable setup processes or options screens, and some of our previous mesh network picks, like the Eero Pro, have emphasized simplicity and ease of use over customizability. But kits with too few settings can also make it hard to troubleshoot problems when trying to set up smart-home devices. The CT8 is just as simple to set up initially as some of our previous picks, but once installed, you have access to more settings that can help you tune the network to your home’s situation. For example, if you find that some devices or laptops in your home keep trying to connect to weaker (but ostensibly “speedy”) signals that provide an unreliable connection, you can set the Asus mesh network to reject weak connections and shunt the devices to the slower but more reliable bands.

The CT8 (and its XT8 brother) also come with subscription-free AiProtection Pro network protection, which includes service optimizations that can prioritize voice and video communications over gaming, among other options. AiProtection Pro’s internet security helps prevent malware from infecting your phones and PCs, and also offers options intended to filter objectionable content from your children’s devices. These are services that would cost extra on the Eero, Motorola, and Netgear Orbi mesh kits.

It only took 10 to 15 minutes from taking the CT8 routers out of the box to connecting laptops to the new network. By default, the network is to exclusively use the second 5 GHz band (hence tri-band) in between the two CT8 routers for speed. If you need (or like) to tweak your network settings for specific cases such as gaming or content streaming, or because you need to give critical devices priority over others, the CT8 offers lots of switches and sliders to change those settings in its administration console.

Close up of the ports on the Asus ZenWiFi AC (CT8).
The CT8 has three LAN Ethernet ports for streaming devices or desktops, plus a USB 3.0 port for network storage. Photo: Michael Murtaugh

It’s physically large, but that gives the CT8 the benefit of extra space for more connection options. Since the router and node are the same, they both have three rear-mounted Ethernet ports and one WAN port (to connect the main unit to your modem). That translates into extra connections for wired devices like streaming boxes, TVs, and game consoles. If you place the mesh node in your den or home office, you can connect your laptop or desktop PC to it for better speeds. It also means that you can connect the second CT8 to the first with an Ethernet cable for quicker speeds between mesh nodes (aka “wired backhaul”), if your home is wired for that.

The CT8 and the XT8 have the ability to rename the 2.4 GHz network separately from the 5 GHz network. This is uncommon among mesh network kits—the Asus ZenWiFi siblings and the Ubiquiti Dream Machine were the only tested mesh networks that can configure separate network names.

Like most mesh networks, the CT8 comes with smart connect on by default. Smart connect (aka bandsteering) works great for laptops and phones: The device in your hands will automatically switch between 5 GHz (short range, but faster) and 2.4 GHz (long range, but slower) as needed throughout the day, using the same network name (SSID, see below). However, some 2.4 GHz smart devices like cameras and smart bulbs have issues with connecting to networks that have one network name. The latter is an ongoing problem with most mesh networks, which lock out settings in the name of simplicity. Changing the network name will alleviate this problem, and you’ll still be able to enjoy the expanded mesh network coverage, though you will lose the ability to roam from 2.4 GHz to 5 GHz automatically on your laptops and phones.

If your household is like mine and you’re adding smart-home devices every year, you’ll need all the help you can get to manage them. Asus routers include Amazon Alexa compatibility, so you can turn on a separate guest Wi-Fi network using your Echo smart speaker, for example. In addition to Alexa support, the CT8 offers the much geekier IFTTT (If This Then That) protocol to automate functions, such as an alert that sends you a check-in email when your kids come home from school and connect their phones to the Wi-Fi.

The Asus ZenWiFi AC (CT8) next to other mesh units to compare size.
The CT8’s mesh units take up a lot more space than rivals. From left: Asus CT8, Eero Pro + Beacon, Trendnet TEW-830MDR2K, Netgear Nighthawk MK62, Motorola MH7023. Photo: Michael Murtaugh

The CT8 mesh units aren’t the largest in our roundup, but they’re close. Asus has made some effort to build as minimalist an exterior as possible, but the CT8 units are a lot larger than the Eero Pro or Netgear Nighthawk MK62 nodes.

The CT8 base unit and satellite look identical, so that could be an issue if you ever move them to another location. Thankfully, when you set them up for the first time, it doesn't matter. You can connect your cable modem or gateway to either node when you’re first installing the mesh, and that unit will act as the base unit after everything is set up. If you ever move things around and try to use the satellite as a router, an error light will alert you to the problem.

WPA3 isn’t available on the Asus ZenWifi CT8. (Though it is available on the XT8, our upgrade pick.) WPA3 is the newest WPA security standard, but lack of WPA3 isn’t a dealbreaker yet, since devices with WPA3 security are still rare. WPA3 will probably be more important in a business environment, like offices and shared hotspots in cafes, in any case.

The CT8 is expandable, but an extra unit goes for about $200 (since they’re the same routers with all the same electronics inside), a significant amount of money for an additional mesh point. Some other mesh kits use proprietary nodes, with different components inside (Google Nest Wifi and Netgear Orbi’s simpler nodes come to mind), which means that expanding the network is often cheaper than with the CT8.

Our upgrade pick the Asus ZenWiFi AX (XT8)
Photo: Michael Murtaugh

Upgrade pick

Asus ZenWiFi AX (XT8)

Asus ZenWiFi AX (XT8)

Ahead of the curve, for gigabit internet

The ZenWiFi AX (XT8) is almost a clone of our CT8 pick, but with significant upgrades in the form of future-proof technologies like Wi-Fi 6 and WPA3. We’d only suggest it if you have a gigabit (1,000 Mbps, or close to it) internet service plan.

Buying Options

$450 from Amazon

May be out of stock

*At the time of publishing, the price was $450.

The Asus WiFi AX (XT8) is the right choice if you currently use a gigabit or faster internet connection (or plan to soon)—otherwise, you may only get minimal benefit for the much higher price. The XT8 is almost a twin of our pick the CT8, but it includes a couple of extras for users who need a more powerful mesh network to work with all that bandwidth. It’s our first pick to support Wi-Fi 6 (instead of Wi-Fi 5, common for the past seven to eight years) and improved security called WPA3. Among other improvements, Wi-Fi 6 should help keep the XT8 relevant in a rapidly changing smart home with more and more devices to manage. The XT8 can also serve a strong, responsive signal to all corners of a larger than average home, and it outperformed the pack in speed and latency in almost all of our tests.

A great router should provide many years of reliable service, and buying one with more recent technology can future-proof it for longer. The Asus ZenWiFi AX (XT8) looks like our top pick, but it adds compatibility for the new Wi-Fi 6 standard, along with extra WPA3 network security. Wi-Fi 6 mesh routers have only been widely available since spring of 2020, are still far more expensive than Wi-Fi 5 routers and don’t bring a huge performance leap to most homes.

The Asus ZenWiFi AX (XT8) next to its AC adapter.
Photo: Michael Murtaugh

The XT8 mesh kit was the top performer in most of our tests, and we think the extra performance will help if you have gigabit internet service; And it was the only mesh kit we tested with a 2.5 Gigabit Ethernet (GbE) port to support faster-than-gigabit connections, if your modem and internet provider supports them. So far, gigabit capable doesn’t necessarily mean that you’ll get gigabit speeds for all of your devices at all spots in your home. The XT8 came the closest to providing this though, achieving a throughput of over 600 Mbps at close range—almost 30% faster than our main pick, the Asus Wi-Fi AC (CT8)

A graph comparing the stacked throughputs of multiple mesh networking kits.
The Asus XT8 topped our throughput chart—it was the fastest mesh kit in most situations. The dual-band Nighthawk MK62 was great near the base unit, but was much worse than the XT8 when communicating through other mesh points. Graph: Wirecutter

When you take raw speed out of the equation, our main pick and upgrade pick are much closer than you’d think in other tests. Their hardware differs slightly, but the CT8 and XT8 use the same Asus routing software, and they performed similarly when all six test laptops were taxing the network at the same time. The XT8’s latency (the time you’d be waiting for sites to respond when browsing them) was only a few percent faster than the CT8. By this measure, XT8 was lockstep with the other top performers. The top four kits use a tri-band network, with a dedicated band for connections between the base unit and satellite(s), which was a main differentiator from kits with more lag.

Close up of the ports of the Asus WiFi AX (XT8).
The XT8’s ports appear almost identical to the CT8, but look closer and you’ll notice the faster 2.5 GHz WAN port and placement of the USB 3.0 port. Photo: Michael Murtaugh

Aside from Wi-Fi 6, WPA3 compatibility, and the speedy 2.5 gigabit port, the XT8 is functionally identical to the CT8. The extra $120 upgrade may make sense for a small slice of people, but we aren’t recommending any of the even more expensive Wi-Fi 6 mesh kits. The next step up from the XT8 is a practically commercial-grade network, where you’ll find features that most people would never touch (and might need an IT degree to understand).

Every time we update our router or mesh-networking guides, readers ask us about enterprise-level networking options like Ubuquiti’s UniFi networking line. Its rack-mounted models are decidedly overkill for most homes, but the more recent UniFi Dream Machine (UDM) seems tailor-made for homes and small businesses. We tested the Wi-Fi 5-compatible UDM in our router guide, but we also wanted to see how it would work with a Ubiquiti mesh extender.

the UniFi AP BeaconHD mesh node
Without accessories like the UniFi AP BeaconHD mesh node, the UDM has a price comparable to that of the Asus ZenWiFi AC (CT8), between $300 and $350. Photo: Ubiquiti

If you link one or more (awkwardly named) Ubiquiti UniFi AP BeaconHD Wi-Fi MeshPoints to the UniFi UDM, you’ll have a mesh network with more expandability and settings to tweak. Pairing the BeaconHD with the UDM has a few more steps than the Asus ZenWifi, but only took a few extra minutes overall. The BeaconHD is a wireless-only extender like Eero’s Beacons, however, so you’ll need another solution if you want a mesh node with an Ethernet port. The BeaconHD improved the Wi-Fi signal to the rest of the house, but as our tests show below, other mesh kits performed a bit better.

A graph comparing the stacked throughputs of multiple mesh networking kits.
The UniFi Dream Machine and BeaconHD provided strong service all over the home, but other mesh kits outperformed it and were easier to set up. Graph: Wirecutter
A bar graph comparing the median latencies of each mesh networking kit.
Latency was a similar story: The UDM and BeaconHD were squarely in the middle of the pack when we tested on a busy network. Graph: Wirecutter

The UDM’s administration app and web interface look polished and professional compared with those of home routers, and they offer plenty of settings and graphic monitors familiar to network engineers. But to folks who just want a simple-to-use router, Ubiquiti’s interface could look like an impenetrable wall of technical details.

Screenshot of a browser showing Ubiquiti’s UniFi controller.
Ubiquiti’s UniFi controller on the UDM shows multiple pages of information about your network and devices. This is great for IT managers, but it can seem like a wall of data.

If you’re an even more technical sort and have a (possibly humongous) home with some truly troublesome coverage zones, you might wonder how a UDM/BeaconHD mess setup compares with a more traditional multiple-access-point setup such as Ubiquiti’s UAP line of devices. Access points like the UAP only transmit Wi-Fi in certain rooms around your home, with wired Ethernet connections back to the router rather than chaining wireless connections between each access point. If you don’t have Ethernet wiring or can’t install it, you can’t use UAP. This arrangement is a much simpler option than mesh, as you don’t have to worry about building materials like stucco blocking a wireless signal between the router and the access points.

The UniFi software on the UDM (or a controlling PC with the app) lets you administer the UAP access points. It’s not that hard to use, but it definitely feels techy enough to scare off some less-technical people.

For the tests in our 2017 house, the UAP-AC-Lite access points were in the same sites as our multi-hop configured mesh kits—one in the network closet, one atop the living room TV island, and one downstairs in the den.

Browsing throughput of Ubiquiti vs Mesh
Graph: Wirecutter

The $80 UAP-AC-Lite units came pretty close to the top of our 2017 charts for throughput, despite that model being the least expensive version in Ubiquiti’s AP line. They didn’t quite take home the gold, with Netgear’s much more expensive Orbi RBK53 kit squeezing out a few Mbps more as tested, but they came close enough that you’d need to squint to tell the difference.

Throughput doesn’t tell the whole story, though. Latency still tends to be a better measure of how your Wi-Fi network feels, and in latency the two UAP units beat the Orbi system (and every other mesh kit we tested in 2017) everywhere.

A graph showing Browsing latency of Ubiquiti vs mesh
Graph: Wirecutter

You’d really feel this difference in heavy use, as the UAP system delivers a smooth and reliable experience with no frustrating pauses or dropped signals. In our tests, it connected rapidly and transferred data smoothly and consistently.

UAP-AC-Lite units currently cost about $80 apiece, but you can use the router you already have on hand. Most of the mesh kits are capable of using a wired connection to any or all of their satellites, but they’re all considerably more expensive than Ubiquiti access points. Even if you add the cost of labor for a professional to run Ethernet cables from, say, your router to the downstairs office and up to a bedroom, the total cost to add three Ubiquiti APs is still currently well under $400, making it a potentially cheaper option for large homes that might otherwise need several mesh nodes to provide reliable service.

Long story short: If you can run Ethernet cables and use standard access points instead of a mesh kit, you should consider it.

With just a few exceptions, testing for most Wi-Fi router reviews consists mostly of connecting a single device to Wi-Fi at various distances, trying to get the biggest throughput number possible and declaring the router with the biggest number and the best range the winner, at least in terms of raw performance. The problem with this method is that it assumes that a big number for one connected device divides evenly into bigger numbers for all devices. This is usually true for wired networking, but it doesn’t work well for Wi-Fi.

Instead of testing for the maximum throughput from a single laptop, we used six, spaced around our test home, in order to simulate the real-world activity of a busy home network. The test home measures over 3,000 square feet, with three floors of living space and a garage with cinder block interior walls.

Because these tests simulated real-world traffic, they did a better job of modeling everyday performance compared with a testing tool like iPerf that simply moves data from one machine to another as fast as possible. We did similar testing for the latest version of our guide to standalone routers.

A floor plan with notes marking where the mesh kits were placed.
These illustrations show where we placed the pieces of each mesh kit in our real-world test environment. Illustration: Sarah MacReading

We verified that the core features of a mesh network were enabled, namely using a single network name (or SSID) to enable roaming for both 2.4 GHz and 5 GHz channels. We also let the router choose its own channels using the automatic setting, if that was the default. For some mesh systems (such as Eero), MU-MIMO is a core feature and cannot be turned off, but if that function was changeable (as in the Orbi system) we left it in its default setting. We didn’t touch most other settings—you should be able to connect to your Wi-Fi and have it work without constantly fiddling with things.

Putting devices in the right places is key to any mesh network’s success. We started by placing the main router or node in the living room, in the center of our testing space, and connecting it to our cable modem via Ethernet.

We placed the second node in the converted attic, which is located on the third floor of the home. It’s about 40 to 45 feet vertically from the base router, through two floors, and at least two interior walls. If present, the third mesh node was placed in the master bedroom on the first floor, with one interior wall between it and the base router. The six laptops (see the diagram above) were placed throughout the home, on all three floors and in the garage situated close to the cinder block foundation.

During testing, the six laptops, our wired controller laptop, and an Apple iPhone running the router app (if needed) were the only devices connected to the test network. We didn’t disable any of the surrounding Wi-Fi networks or wireless devices like Google Home speakers; these kept doing their usual noisy things, just as they probably do in your home. The neighbors and our home network also kept their Wi-Fi networks going, which left somewhere in the vicinity of a half-dozen to a dozen network names visible at any given time.

Our six laptops ran the following tests:

  • Browsers: Three laptops simulated real-human web browsing by loading a “web page” once every 20 seconds. Each “web page” consisted of 16 separate 128 KB files, all requested simultaneously, and we measured latency from the time the requests went out to the time all 16 requests were fulfilled. This is the most important test—it accurately represents the thing that frustrates real users most (slow and inconsistent web browsing)—and it usually fails before any of the other tests do.
  • Downloader: One laptop downloaded a very large file.1 We wanted to see an overall throughput of 100 Mbps or better, to simulate the experience of an impatient person waiting for a device to complete an update. This test is a big challenge for the rest of the network—if this laptop gets all of the available airtime, the other tests suffer.
  • Video streamers: Two laptops each simulated a 4K video streaming session. They tried to download data at up to 30 Mbps, but we were satisfied if they could average 25 Mbps or better, which is what Netflix recommends for 4K. If this laptop can’t get at least 20 to 25 Mbps, that means a real video would be pausing and buffering. Like the test involving the download laptop, this test presents a real challenge to the rest of the network.

These tests simultaneously evaluated range, throughput, and the router’s ability to multitask. We ran all these tests at the same time for a full five minutes to simulate a realistic extra-busy time on a home network. Although your network probably isn’t always that busy, those busy times are when you’re most likely to get frustrated by poor network performance. We ran each test six times, and we then averaged the results.

We also tested raw speed in terms of throughput at the farthest spot in the attic/sunroom, at a closer spot about 15 feet away with no obstructions, and in a bedroom on the second floor with several interior walls in between the laptop and the main router. We wanted to make it difficult for the mesh-router kits but not impossible, so we surveyed and dismissed other spots in the home during our initial setup.

Testing mesh kits this way—with a mix of easy and difficult spots to reach—ensures that we find the ones that work best throughout your house, rather than just looking good in the easy spots.

In addition to testing for raw throughput and the quality of web browsing, we made sure roaming worked well on our picks by checking each router’s interface (if present) to confirm that all the laptops weren’t bunched on a single node, router, or satellite.

We believe that mesh’s purpose isn’t to make Wi-Fi fast somewhere—it’s to make Wi-Fi fast everywhere. So we tested throughput in three spots in our test space, with increasing degrees of difficulty: A spot in the living room, with a clear 15 foot run to the mesh base unit, in the attic, where the test laptop would be best situated to connect to the satellite node there, and a spot in one of the home’s bedrooms, where the laptop could choose to connect to a satellite or through several walls to the main router.

A graph comparing the stacked throughputs of multiple mesh networking kits.
For the throughput test, we had each laptop download the same 16 MB file repeatedly for 30 seconds. We measured the results in megabits per second (Mbps), and higher numbers are better. Graph: Wirecutter

As the throughput graph above shows, the mesh kits we tested with an efficient wireless backhaul trounced our budget standalone router pick, the TP-Link Archer A7, across the board. The better-performing mesh kits, including the tri-band Asus ZenWifi CT8, our pick, were able to handle each situation adroitly and provided over 400 Mbps at close range and over 100 Mbps in the other situations.

This efficient relay system is the hallmark of a good mesh network. You can see that in the “one jump” results in the attic. Good tri-band kits like the Asus ZenWifi twins and the D-Link COVR-2202 utilized their dedicated backhaul to relay network signals efficiently to the base station. Dual-band kits like the Netgear Nighthawk MK62 took fairly significant hits to performance, as the 5 GHz radios in those kits have to split time between connecting local clients like the laptops and communication with the base units.

A floor plan with notes marking where the mesh kits were placed.
These illustrations show where we placed the pieces of each mesh kit in our real-world test environment. Illustration: Sarah MacReading

That throughput graph is worth looking at to get an idea of your best-case performance when you’re the only one on the network, but those test results don’t tell the most important story. To do that, we needed multiple laptops, simulating a busy little real-world network; you can read more about how we set up the network and what tests we ran above. A brief recap: We had two laptops simulating a 4K video screen, one downloading large files, and three laptops browsing the web.

The web browsing test is both the most realistic representation of your experience in using your Wi-Fi and the test that almost always fails before any other test does. By running multiple web browsing tests while other laptops are streaming video and large files, we test the worst-case scenario for a mesh-networking kit.

Modern web pages consist of a complex set of files that must all be fetched before the page renders in your browser. A Facebook feed, for example, has multiple resources, like HTML code, CSS libraries, and Javascript, that load after each click. An error in one of those resources can block the entire page from displaying in your browser, slowing your experience. As a result, a simple task—click a link, get a web page—is actually complex, and a relatively minor number of errors and slowdowns can magnify rapidly into a situation where web pages don’t load and you have to hit the refresh button. Our web browsing test, by more closely simulating what your browser really has to do behind the scenes, exposes such problems more accurately.

Check out this graph, which gives you a sense of how long you might have to wait around for things to load:

A bar graph comparing the median latencies of each mesh networking kit.
This chart shows the median latency of each workload and site we tested during our multiple-client tests. The most important figures here are the Web browsing latency readings: A big browsing bar means frustration waiting for slow page loads. Graph: Wirecutter

For web browsing, latency—how long it takes between a request and a response—is more important than raw throughput. By looking at the latency of requests made while all six laptops were working simultaneously, we got a good measure of how well the network might function for you during a busy time.

The best mesh-networking kits occupied similar spots in the stacked latency chart compared with the throughput charts, though there was some movement in the middle and bottom of the pack. Significantly, the Netgear Nighthawk MK62 dropped from sixth overall on the throughput chart down to 13th overall on the stacked latency chart. The browsing streams on the MK62 have longer lines, showing that the dual-band mesh network took more time to juggle between the clients and the nodes.

The next two charts show what percentage of the time our test networks delivered a satisfactorily fast browsing experience. While many of the tested networks dragged 5% to 10% of the time (the spike on the far right), our top performers handled the vast majority of these tests in stride.

Rather than just looking at the mean (average) of all the results during our five-minute test window, we look across the spectrum of results, from good to bad. The median—or typical—result is the 50th percentile, on the left. Moving right, the 75th percentile shows you how bad one out of every four clicks will be, and the 99th/100th percentile is the worst result of the five-minute run. If a kit rockets off the chart at the 75th percentile, for example, that means one frustratingly slow page loads out of every four. This is a tough test, but we think a good network should respond reliably all the time.

A graph comparing the web latency performances of multiple mesh networking kits.
In this test we looked at the time it took to fetch our 16-element “web page,” by percentile. The median (50th percentile) results are on the left, while the worst (99th percentile) results on the right; lower results are better. All four of the top performers were able to handle these tests in stride, returning a satisfactory browsing experience throughout each test. Graph: Wirecutter
A graph comparing the web latency performances of multiple mesh networking kits.
The rest of the pack had more problems with latency, though most were able to outperform the standalone TP-Link Archer A7 router we used as a control. Graph: Wirecutter

In the graphs above, we show the topmost and lowest performance groups individually, as measured by the average time to load web pages to the three browsing laptops during our multi-client testing. Most of the top performers have tri-band radios; these results show the difference that extra radio makes while the network is transferring data to the satellite(s) and therefore to the more distant clients. This effect shows up best when you’re comparing a strong tri-band kit like the Asus ZenWiFi AC (CT8) set or Eero Pro 6 with a kit such as the TP-Link Deco X20: The tri-band CT8 was rock steady in the green throughout our test suite, while the dual-band Deco X20 started to falter after the 80th percentile. Don’t count dual-band kits out entirely, however: The Ubiquiti UDM with BeaconHD was competitive on this test, while the tri-band Trendnet TEW-830 was overwhelmed relatively quickly.

We wanted to test an inexpensive mesh kit like the TP-Link W2400 and compare it with the other mesh networks here. While it placed near the bottom on our close-distance throughput tests, it was still faster than the Archer A7 standalone router we used for comparison at the closest distance. The Deco W2400 also showed better performance on the latency tests, remaining competitive with much more expensive kits like the Eero Pro with 2 Beacons and the Ubiquiti Dream Machine with UniFi BeaconHD. It did fail one of our 4K streaming tests and one of the browsing tests when the network was fully loaded, so it couldn’t replace our budget pick.

A mesh network extends Wi-Fi to all corners of your home by using multiple plug-in boxes generically called mesh nodes. Whether the nodes are called Beacons (Eero and Ubiquiti), Deco units (TP-Link Deco), or Satellites (Netgear Orbi), they essentially do the same thing: pass and repeat Wi-Fi around signal–blocking materials such as masonry walls or metal doors, or bring Wi-Fi service to parts of your home that are out of range of a single standalone router.

Here are a few terms that describe the various parts of a mesh network:

Router or base unit: This is the device you set up first. It connects to your home’s internet (via an Ethernet connection to a cable modem or the gateway router) and broadcasts Wi-Fi.

Mesh node or satellite: These are the devices that connect back to your router (or another node) to extend that network and provide a more reliable Wi-Fi connection over a greater area. Most kits come with one or two of them. Netgear calls them Orbi Satellites, Eero calls them Beacons, and D-Link calls them Covr Points. Sometimes they are physically identical to the base unit, sometimes they aren’t.

Access point: An access point provides a Wi-Fi connection to your devices but passes all its data back to the main router via an Ethernet cable to be sorted. If Ethernet is an option in your home, you should use wired access points rather than mesh.

Wi-Fi range extender: Wi-Fi range extenders (also known as Wi-Fi signal boosters) are less expensive than mesh nodes but are also slower and less capable. A range extender will generally create a second network name (SSID) when you set it up, and you’ll need to change networks on your phone or other devices manually when you switch from the standalone main router to the extender in your home.

SSID or network name: An SSID is the fancy term for a Wi-Fi network’’s name.

Bands and channels: A dual-band mesh kit communicates with devices on two sets of radio frequencies (aka bands), 2.4 GHz and 5 GHz, while a tri-band kit has an extra 5 GHz band that can help with communication between the router and satellites. Each band is broken down into separate channels so more devices and networks can communicate in the same area without interfering with each other.

2.4 GHz versus 5 GHz: The 2.4 GHz band is slower but is compatible with more devices and can reach farther and through walls better. The 5 GHz band is faster but has a shorter range.

Dedicated backhaul: Dedicated wireless backhaul is a wireless band that serves only for communication between the router and its nodes, not the connection to computers, phones, or other devices. Some mesh networks can use Ethernet wires as a backhaul, which is even faster.

MU-MIMO: This term stands for Multiple User, Multiple Input, Multiple Output. If routers and devices both support it, MIMO can speed up Wi-Fi connections. With MIMO, each band has multiple data streams that can work simultaneously between the device and the router, similar to adding lanes to a road. Before MU-MIMO, it could talk to only one device at a time. (If you’re wondering how routers have been managing multiple Wi-Fi devices up till now if they can actually talk to just one device at a time, it’s by switching between devices very quickly.)

Wi-Fi 6 (aka 802.11ax): The 802.11ax protocol, also known as Wi-Fi 6, will replace the current 802.11ac (Wi-Fi 5) protocol in over the next year or two the same way 802.11ac replaced 802.11n nearly a decade ago. Some improvements will help with overall speed, but we’re most interested in improvements like MU-MIMO and OFDMA; Clunky acronyms that ultimately should make Wi-Fi better at managing busy home networks full of computers, phones, streaming boxes, smart devices, and the like. These technologies tout new capabilities to help avoid interference in dense areas where neighboring networks fight one another.

New terminology defining an expansion of the Wi-Fi 6 protocol, called Wi-Fi 6E, has been announced. This expanded wireless standard will take advantage of newly deregulated 6 GHz radio frequencies. The new frequencies will reportedly let you add up to 65 devices to your mesh network, while Wi-Fi 5-only networks can start to lag when you approach 30-35 devices.

At this year’s virtual CES 2021, Wi-Fi 6E mesh networks with 6 GHz channels included Linksys’ AXE8400, the TP-Link Deco X76 Plus tri-band mesh kit, and TP-Link Deco X96. Wi-Fi 6E clients (phones, laptops, etc.) using the 6 GHz radio channels are going to be very rare for a few years, so these expensive mesh kits may be hard to recommend for most people.

Wi-Fi 6 (sans ‘E’) mesh networks will continue to be released, and will be our focus for 2021. TP-Link also released the Deco Voice X20 with built-in Alexa-compatible speakers (like the Google Nest Wi-Fi), Arris promoted the SURFBoard mAX AX6600, and D-Link released the AI M32 AX3200 mesh kit.

TP-Link announced the Deco X90 mesh system at CES 2020, but it still has not been released.

Amplifi released its Alien mesh kit with Wi-Fi 6 and an innovative touchscreen control panel. However, it’s not compatible with current Wi-Fi 5 Amplifi kits (see below), and its $700 price tag is higher than the pricier tri-band kits we’ve tested recently.

Arris has released the SURFboard mAX mesh Wi-Fi system based around three models with increasing speed promises (and prices). We’re considering these kits for our next guide update.

Our previous pick: Eero Pro + 2 Eero Beacons
In the previous version of this guide, we praised the Wi-Fi 5 Eero Pro + 2 Eero Beacons kit’s ease of setup and its performance on our latency tests. However, the competition has surpassed it in price and performance on our tests.

Synology RT2600ac and MR2200ac
The Synology mesh routers are still a good combination, if you’re into tuning your network for problem devices. For example, like the Asus ZenWiFi twins, you can separate the Synology’s 2.4 GHz band into a separate SSID to facilitate connecting smart devices like cameras and smart switches.

D-Link COVR-2202-US
The D-Link COVR-2202-US was our previous budget pick, but D-Link has told us that it will be replaced in their lineup with a selection of mesh-capable routers and range extenders. The COVR-2202 did well in our tests against the new contenders, but we have also read reports that some folks have had technical issues with the COVR mesh kit. D-Link doesn’t sell individual COVR-2200 nodes, so the kit has limited expandiability, beyond buying multiple kits. As of this writing, COVR-2202 stocks are low, so it shouldn’t be too hard to avoid it.

Eero 6
The Wi-Fi 6 version of Eero’s basic mesh kit has an attractive price, particularly when it's on sale, and a recent firmware update has improved its performance on our tests. However, our pick the Asus ZenWiFi CT8 is still faster and a better buy because the Asus includes parental controls and internet security; features that Eero locks behind a monthly subscription.

The Eero 6 is very simple to set up, so it would be a natural fit for folks who want to minimize their time fiddling with router settings. All Eero and Eero Pro kits now have a temporary off switch for its 5 GHz network, to make it easier to connect smart devices like cameras and doorbells. It has support for WPA3 security, but that’s one of the few settings you can change beyond network name and password. In its quest for simplicity, it removes even basic niceties (it has only one extra Ethernet port on the base unit, and none on the satellites). Eero offers a pricier package that adds four additional Ethernet ports for about $70.

During our testing, the Eero 6’s performance was competitive when the laptop was connected to the base router, but throughput dropped off precipitously when the signal was bounced through the satellites. That means it is suitable for homes with 500 Mbps (or slower) broadband connections. Eero also charges between $30 and $120 per year for their Eero Secure and Secure+ subscriptions, which are necessary for parental controls and anti-malware protection—services that the Asus ZenWiFi kits include for free. This makes the budget-priced Eero 6 less of a bargain.

Eero Pro 6
Like its budget-priced cousin, the Eero Pro 6 is easy to set up, and removes many router settings for simplicity’s sake. However, as a three-pack of identical Wi-Fi 6 routers, it’s one of the most expensive mesh kits we’ve tested so far, almost double the price of the Asus ZenWiFi CT8, our pick.

Eero Pro 6 has two Ethernet ports on each router, which will allow Ethernet backhaul connections between the routers, as well as local wired connections. They’re also tri-band routers, with latency and throughput performance that is competitive with the Asus ZenWiFi models. Like other Eero kits, you have to register the Eero Pro 6 online with Amazon and administer the mesh network in the cloud. Technically, bad actors could monitor your data via this login, though Amazon and Eero assure us that they safeguard your data and online activities. Like the other Eero kits here, online security and parental controls require a $30-120 per year subscription. The Eero Pro 6 is a solid alternative as a simple, set-it-and-forget-it networking solution, if you can absorb the monetary costs. Ultimately, we prefer the Asus WiFi AC and AX over the Eero Pro 6 due to the latter’s high purchase price and expensive extras.

Linksys Velop AX4200 (MX12600)
The Linksys Velop AX4200 (MX12600) is another tri-band kit that shows excellent performance on our latency and throughput tests, finishing just behind the Asus ZenWiFi AC, ZenWiFi AX, and the Eero Pro 6. However, like the Eero Pro 6, the MX12600 is pricey—currently $50-170 more expensive than the Asus kits. The MX12600 only has basic parental controls and lacks malware protections beyond what you install on your devices, while the ZenWiFi kits have robust security and parental controls built-in. Last, but not least, Linksys also requires an online login during setup: You could use a burner email account to protect against online threats, so it’s arguably more secure than using your Amazon login on the Eero kits.

Motorola MH7023
The MH7023 did quite well on our latency tests, though it was middle of the pack on overall throughput tests. The MH7023 is a new 2020 mesh kit, and as Motorola’s first mesh kit, it doesn’t have the track record that other network makers have. We’ll be looking out for reviews of the MH7023 and see if others have good experiences with it as well. We may reevaluate it in the future.

Netgear Nighthawk MK62
The dual-band Netgear Nighthawk MK62 (and its three-piece sibling, the MK63) is a relatively inexpensive mesh kit, with small nodes that are easy to hide. It did reasonably well on our throughput tests, but it was less able to handle multiple clients on our latency tests, placing close to last.

Ubiquiti UniFi Dream Machine + BeaconHD
We’ve written a bunch more about the Ubiquiti Dream Machine (UDM) here and in our standalone router guide, but the short story is that the UDM with BeaconHD is a great entry point into the world of enterprise-level wireless networking. It’s something you’d want to look into if you really want to spend hours tweaking, have an Ethernet wired home, or can spend a lot of time adding UniFi access points to create a bespoke solution for your (hopefully large) home.

Trendnet TEW-830MDR2K
The Trendnet TEW-830MDR2K was one of the least expensive mesh kits in our roundup and also had small unobtrusive nodes. Each node had screw mounting holes molded into their bases, which is a nice touch. However, the Trendnet’s admin interface was unpolished, and it placed dead last on our latency tests.

TP-Link Deco X20
The Deco X20 was middle of the pack in price and performance, but it failed one of our 4K streams (way under 20 Mbps) when the network was fully active.

TP-Link Deco W2400
The Deco W2400 is a Walmart exclusive, but we wanted to test it because it is one of the least expensive mesh-network kits on the market, under $100. It was the slowest performer on our throughput tests, but it surprised us with upper middle-of-the-pack latency results. It did fail one of our 4K streaming tests and one of the browsing tests when the network was fully loaded. This kit might work if you want to inexpensively outfit a spread-out home that doesn’t have a lot of devices and has a basic internet plan (under 100 Mbit).

Netgear Orbi RBK752
In addition to being one of the most expensive mesh kits here, the Orbi RBK752 failed one of our 4K streaming tests while the network was crowded. We also noted complaints in our comments and in Amazon reviews about Netgear’s short, 90-day phone tech support policy, which has complicated buyers’ ability to obtain service during the one-year warranty period.

We evaluated and passed on expensive mesh networking kits like the AmpliFi Alien Mesh, Arris SURFboard mAX Pro (AX11000), Linksys Velop Wi-Fi 6 (MX10), Netgear Orbi852/RBK853, and TP-Link Deco X60. These kits promised more performance, added range, extra capacity, and more powerful radios. We dismissed them because they seemed to be overkill and were expensive compared to the tri-band and dual-band competitors above. We also considered the Asus RT-AX92U mesh kit, but passed, because it uses its Wi-Fi 6 radios solely for wireless backhaul when setup as a mesh network.

Our previous pick: Netgear Orbi RBK50
Netgear’s Orbi RBK50 was our pick in the 2018 iteration of this mesh-networking guide. It’s been supplanted by the RBK752 and RBK852 in Netgear’s Orbi lineup, but the RBK50 is still available at a lower price. The RBK50 is one of the first successful mesh kits, but we think our current picks have surpassed it.

Google Nest Wifi
The Google Nest Wifi is available as a 2-pack or a 3-pack. The Nest Wifi router is a dedicated Wi-Fi 5 base unit with two Ethernet ports, and the wireless-only Nest Wifi access points have Google Assistant functionality and a smart speaker built in. In our testing, the Nest Wifi kit excelled at the single-client throughput tests but lagged the Eero Pro + 2 Beacons when all six laptops were using the network at the same time. While the Nest Wifi is easy to set up if you use Google’s services often, there are privacy concerns with Google’s Assistant technology, which extend to the websites and services you visit using the Nest Wifi.

Linksys Velop Tri-Band (WHW0303) and Linksys Velop Dual-Band (WHW0103)
We evaluated the tri-band Linksys Velop WHW0303 and dual-band version of the Linksys Velop router kit in a previous version of our guide. Neither kit was a top pick. The tri-band kit was a strong competitor, but it was (and still is) too expensive.

Eero 3-pack
The Wi-Fi 5 version of the Eero mesh kit is compatible with previous Eero hardware, so you can use them to expand an existing Eero mesh network. However, in previous testing, the three-piece Eero kit failed one of our two 4K streaming tests, while its other performance was middling.

D-Link COVR-C1203
The D-Link COVR-C1203 is a dual-band three-piece mesh-networking kit that is easy to hide virtually anywhere in your home. However, you’ll have to buy a second three-piece kit to expand the mesh network in the future. The COVR-C1203 also failed one of its two 4K streams during 2019 testing, and its overall performance ranked at or close to the bottom of the field.

TP-Link Deco M4, Deco M9 Plus, and Netgear Orbi RBK13
The TP-Link Deco M4 and Deco M9 Plus are compatible with other Deco kits, and the Deco M9 Plus has a smart hub built in for Zigbee and Bluetooth devices, but most folks don’t need a smart hub. Netgear’s Orbi RBK13 is a dual-band, three-piece mesh-networking kit featuring a router and satellites that are the same size as the Netgear Nighthawk MK62. Its tested performance in 2019 was similar to the TP-Link Deco M4 and Deco M9 Plus: middling.

AmpliFi Instant
Like the AmpliFi HD, which we tested and dismissed in 2017, the compact and inexpensive AmpliFi Instant has a front-mounted touchscreen that lets you monitor your network bandwidth. We found it easy to set up, and it’s compatible with AmpliFi HD routers and mesh satellites. However, it didn’t impress us much—it underperformed compared with other mesh kits and even with our standalone router pick.

Plume and Samsung SmartThings Wifi
Plume’s pitch sounds great: a mesh network that actively adapts to your network needs and any new devices that connect. Plume’s prime selling point is also its greatest drawback: you are locked into paying for an ongoing $99 annual subscription for cloud-based, active network optimization. Since we’ve started following Plume, they’ve expanded their purview to providing devices and services to ISPs like Comcast and Bell Canada. We prefer hardware that you own outright, and doesn’t require paying for a subscription.

We tested the Samsung SmartThings Wifi kit in a previous version of this guide. It has a lifetime subscription to Plume, and has a built-in SmartThings hub, but it finished close to the bottom of the pack in 2019.

Zyxel Multy U
The Zyxel Multy U is a tri-band three-piece mesh-networking kit designed to look good. However, its poor browsing performance in our tests derailed any appreciation we had for its aesthetics.

The rest (2017–2018 testing)

Over the years, we’ve tested the Google Wifi, Netgear Orbi RBK53, Orbi CBK40, RBK40, RBK30, TP-Link’s Deco M5, Eero Pro 3-pack, AmpliFi HD, Amped Ally, Luma, and Zyxel Multy X mesh networking kits. We’ve found them all wanting for one reason or another, and dismissed them from this guide update.

  1. The test procedure actually downloaded a 1 MB file over and over as fast as it could. This test measured throughput and speed, and—more important—put a constant load on the router and satellite(s) that the other devices on the network had to work around.

    Jump back.
  1. Dong Ngo, Eero Wi-Fi System review, CNET, February 23, 2016

  2. Jim Salter, 802.eleventy what? A deep dive into why Wi-Fi kind of sucks, Ars Technica, March 4, 2017

  3. Dan Goodin, Serious flaws leave WPA3 vulnerable to hacks that steal Wi-Fi passwords, Ars Technica, April 11, 2019

  4. Eric Geier, What is WPA3? And some gotchas to watch out for in this Wi-Fi security upgrade, Network World, November 2, 2018

  5. Dong Ngo, Asus ZenWiFi AX Review: The Best AiMesh System to Date, Dong Knows Tech, February 4, 2020

About your guide

Joel Santo Domingo

Joel Santo Domingo

Joel Santo Domingo is a senior staff writer covering networking and storage at Wirecutter. Previously he tested and reviewed more than a thousand PCs and tech devices for PCMag and other sites over 17 years. Joel became attracted to service journalism after answering many “What’s good?” questions while working as an IT manager and technician.

Edit
Dismiss