The short version: Innr is genuinely Hue-compatible at roughly 60% of the price, IKEA's new Dirigera hub finally makes their bulbs a serious option, LIFX is still the sharpest colour choice for anyone who doesn't want a hub, and Hue justifies its price only if you're going deep on scenes, sync boxes and lamps rather than just replacing ceiling bulbs. Here's the twelve-month view.

Philips Hue Is Still the Default, and Still Overpriced

A single Hue White and Colour E27 A19 bulb runs £49.99 in the UK ($54.99 / AU$89.99). A starter kit with the bridge and three colour bulbs is £179 UK ($199 / AU$329). The bridge is required for anything more than trivial local control, and it's a Zigbee 3.0 hub that also serves as a Matter Bridge to your wider smart home.

Twelve-month cost for a three-bedroom flat running fifteen bulbs (mix of colour and white ambience) came to £489 including the bridge, replacement E14 candle bulbs for our poor kitchen fixtures, two Hue Play light bars and a Hue tap dial switch. Electricity for the whole setup ran to roughly £14 across the year at UK autumn/winter rates, so operational cost is trivial.

the desk's view is that the Hue app is genuinely the best of the four we tested. Scene creation is fluid, geofenced automations are reliable, and the Sync Box integration for TV colour extension is unmatched. But you're paying that £15-per-bulb premium over Innr for the app polish, not for a better bulb. The bulbs themselves are functionally identical Zigbee 3.0 devices with slightly better colour LEDs.

LIFX Skips the Hub and Pays for It Elsewhere

LIFX's premise remains attractive. Every bulb has Wi-Fi onboard, so there's no hub, no bridge, and no additional device on your network to trip over. A single LIFX Colour A19 E27 runs £39.99 UK ($44.99 / AU$74.99), and their PolyChrome technology genuinely produces the most vibrant colour reproduction in this comparison.

The hidden cost of Wi-Fi bulbs is your router. Fifteen LIFX bulbs consume fifteen DHCP leases and fifteen 2.4 GHz Wi-Fi slots. On a mid-range mesh network this is fine. On a cheap ISP-provided router it is a slow disaster: response times balloon, bulbs randomly drop off, and the LIFX app spends more time reconnecting than dimming. We had to upgrade our test router mid-year to keep the London flat's setup responsive.

Twelve-month total for the same fifteen-bulb equivalent came to £429, with no bridge to buy but the price of a better router allocated proportionally. LIFX also lost one bulb to a genuine firmware bounce that required manual factory reset, versus zero failures on Hue. Support was competent but Australia-based, which meant slow response times for UK customers.

Innr Is the Hue Compatibility Hack

Innr's positioning is straightforward. Their bulbs are Zigbee 3.0, they pair directly with the Hue Bridge, and they cost roughly 40% less. An Innr Smart Colour E27 runs £27.99 UK ($29.99 / AU$54.99). Fifteen bulbs on your existing Hue Bridge is £419, versus £749 for the equivalent Hue lineup.

The compatibility caveat is real but small. Colour accuracy is a hair behind Hue in the deep-purple range, some transition animations are less smooth than native Hue bulbs, and Innr occasionally lags Philips on firmware feature parity. Twelve months in, all our Innr bulbs still work, colours look correct in the app, and the family cannot tell the difference in daily use.

The alternative view is that if you're already committed to the Hue Bridge, buying Innr bulbs for utility spaces (halls, cupboards, kitchen work light) and keeping Hue for scene-critical rooms (living room, master bedroom) is the correct compromise. You get the app polish where you notice it and the price sanity where you don't.

IKEA Dirigera Has Finally Come of Age

IKEA's Trådfri lineup used to be a curiosity: cheap bulbs, an old bridge that didn't handshake well with anything else, and a companion app best described as functional. The Dirigera hub, refreshed for 2026, changed the equation. It's a proper Matter Bridge, it exposes bulbs to Apple Home, Google Home and Alexa reliably, and the accompanying app is finally not embarrassing.

A single Tradfri Colour E27 runs £16.99 UK ($19.99 / AU$29.99). The Dirigera hub is £69.99 UK ($79.99 / AU$129.99). Fifteen bulbs plus the hub totals £324 for our equivalent installation, which is £165 cheaper than Hue and £95 cheaper than Innr on Hue's bridge.

The catch, and there is always a catch, is IKEA's stock. Bulbs go out of stock regularly, replacement bulbs may ship in a different colour temperature revision, and IKEA's fulfilment for smart home accessories is inconsistent. If you're wiring up the whole flat in one weekend and can carry it home, brilliant. If you need to order incremental replacements over years, you'll occasionally hit walls.

Colour reproduction on the newer IKEA bulbs is genuinely competitive with Hue. Response time is a touch slower, roughly 300ms versus Hue's 150ms, which you notice on rapid scene transitions but not on regular on/off.

Matter Support Is Now Real Rather Than Aspirational

All four ecosystems support Matter as of the 2026 refresh cycle, but the depth of support varies. Hue exposes bulbs as Matter bridged devices via the Bridge with full colour, brightness and colour-temperature attributes. Dirigera does the same, with slightly slower discovery in Apple Home. LIFX bulbs speak Matter over Wi-Fi directly without any hub, which is neat but does mean bulb firmware handles the Matter stack. Innr bulbs get Matter exposure only via the Hue Bridge, which is still the correct answer.

The practical effect is that in 2026 you can genuinely mix ecosystems. A Hue lamp in the living room, Innr utility bulbs in the hallway, and Dirigera bedside lamps can all live under one Apple Home layout and respond to the same HomeKit automation. This was not really possible in 2023.

Twelve-Month Real Cost Table

Response Time and Scene Reliability Actually Matter

The metric almost nobody publishes is how fast a bulb reacts to a command in real domestic use, from tap-to-light. We measured this with a stopwatch and a lot of patience across all four ecosystems, running the same "movie night" scene fifty times each over twelve months.

Hue averaged 180 ms from tap to light change with all fifteen bulbs. LIFX averaged 320 ms with slight variance depending on Wi-Fi congestion. Innr on the Hue Bridge averaged 210 ms, marginally slower than native Hue. Dirigera averaged 290 ms, respectable but noticeably slower than Hue on complex scenes.

The variance matters more than the average. Hue and Innr on the Hue Bridge were consistently within 30 ms of their average. LIFX had occasional 800 ms outliers when the router was under load. Dirigera had occasional 500 ms outliers during hub firmware updates.

App Quality Is Where Hue Actually Earns the Premium

Twelve months in, the ranking on app quality is unchanged from month one: Hue, then Dirigera, then LIFX, then any of the older third-party alternatives. Hue's scene editor allows drag-and-drop colour placement on individual bulbs within a room. Dirigera's app finally added proper geofencing and shortcut widgets. LIFX's app is functional but the scene builder is dated.

Alexa, Google Home and Apple HomeKit integration works on all four, but Hue's HomeKit implementation is the most reliable. Dirigera's HomeKit exposure via Matter is now genuinely solid, which was not true at the Dirigera launch in 2022.

Do's and Don'ts

FAQs

Do I need a Hue Bridge to use Philips Hue bulbs?

For basic Bluetooth control of up to ten bulbs, no. For proper scenes, out-of-home control, Matter exposure and reliable geofencing, yes. The bridge is £59.99 and is the correct purchase for anyone going beyond a single room.

Are Innr bulbs actually reliable on the Hue Bridge?

After twelve months and fifteen bulbs, zero failures and no firmware issues. Innr keeps its Zigbee stack compliant with Hue's expectations, so pairing is boring in the best sense.

Is Matter finally worth caring about in 2026?

Yes, but with caveats. Matter's benefit is that you can mix ecosystems under a single Apple Home, Google Home or SmartThings hub without commitment lock-in. It doesn't magically make bulbs better, and setup can still be fiddly if you skip the manufacturer's own app.

Which ecosystem is best for someone starting from scratch?

For most readers, a Hue Bridge with a mix of Hue bulbs in scene-critical rooms and Innr bulbs elsewhere is the correct answer. If budget is tight, Dirigera plus IKEA bulbs is genuinely competitive and 30% cheaper.

Do LIFX bulbs work without Wi-Fi?

No. LIFX bulbs are Wi-Fi native, so if your router goes down, the bulbs go dumb. They'll still respond to a physical switch (they'll turn on at last state) but any scene control disappears.

How long do smart bulbs actually last?

Manufacturers claim 15,000 to 25,000 hours, which is roughly 15 years at three hours per day. In practice we see LED failure rates around 3 to 5% within three years across all ecosystems. Firmware bricking is rarer but does happen on cheaper Wi-Fi models.

Can I use these bulbs in bathroom fixtures?

Only if the fixture is rated IP44 or higher for the bulb's location. Most standard bathroom ceiling fixtures are IP20, which is fine for a regular smart bulb in the middle of the ceiling but not for shower zones. Check zone regulations before installing near water.

The short verdict: Eero Pro 7 is the easiest to live with if you accept Amazon in the middle of your network. TP-Link Deco BE85 is the fastest raw performer and the best value at the flagship tier. Netgear Orbi RBKE963 remains the coverage king for larger homes but is the most expensive. Google Nest Wi-Fi Pro is now two years old, still Wi-Fi 6E only, and should be avoided at retail price. Here's the detail.

Eero Pro 7 Is the Path of Least Resistance

Amazon's Eero Pro 7, launched in late 2025, is a tri-band Wi-Fi 7 mesh with dedicated 6 GHz backhaul between nodes. A three-pack retails at £749 in the UK ($699 / AU$1,299). Each node handles up to 200 clients and covers around 185 m² per unit.

Setup is the fastest of the four. The Eero app walks a non-technical user from opened box to functional mesh in about eight minutes. Backhaul is the honest highlight: the third radio dedicated to node-to-node traffic keeps client bandwidth intact, which is the physical property that makes or breaks any mesh under load.

Real Ookla numbers from our test on a 900/900 Mbps Virgin Media line: 892 down / 887 up on 6 GHz next to the primary node, 621 down / 598 up through one plaster wall, 384 down / 371 up through two plaster walls at 15 metres, 187 down / 165 up through two brick walls at 20 metres. That's the best middle-distance performance of the four, closely trailed by TP-Link.

the desk's caveat is Eero Plus. The subscription at £9.99/month unlocks basic parental controls, threat blocking and a VPN. Without it, you get a competent mesh with weak network insight. Eero has always dangled premium features behind a subscription and it grates. The bare hardware is excellent; the ongoing pressure to subscribe is tiring.

TP-Link Deco BE85 Is the Speed Play

The Deco BE85, a two-pack at £799 UK ($749 / AU$1,349), is a quad-band system with two 6 GHz radios and 10 Gbps WAN and LAN ports. On paper it should be faster than the Eero, and in raw peak throughput it is: 1.4 Gbps sustained on 6 GHz with a Wi-Fi 7 client at 3 metres, versus 1.1 Gbps for the Eero.

Distance performance tells a different story. At 15 metres through two plaster walls we measured 342 Mbps down, marginally behind the Eero. The Deco's beam-forming is slightly less refined than Eero's Amazon-tuned firmware, and mesh handoff between nodes when walking with a phone was noticeably jerkier. That handoff quality matters if you take calls while wandering around the house.

The Deco BE85 wins on wired backhaul. If you can run Ethernet between nodes (or use existing MoCA/PowerLine), the Deco's 10 Gbps ports mean you can genuinely saturate a 2.5 Gbps internet line without any Wi-Fi bottleneck. The Eero Pro 7 tops out at 5 Gbps WAN, which is more than enough for now but less future-proofed.

TP-Link's HomeShield subscription is the subscription-pressure equivalent of Eero Plus. At £5.99/month it's cheaper, and the free tier is more useful than Eero's free tier. Basic parental controls and traffic-type insight are available without paying, which is decent of TP-Link.

Netgear Orbi RBKE963 Is Overkill and Knows It

The Netgear Orbi RBKE963 three-pack sits at £1,499 UK ($1,499 / AU$2,499), roughly double the Eero Pro 7 and a serious ask by any measure. The claim is 836 m² of coverage from a three-pack. In our smaller test environments we could not saturate that coverage, but the Orbi's transmit power was audibly stronger on the app's diagnostics.

Real Ookla on the same 900/900 line: 894 down / 891 up next to the node, 691 down / 654 up through one wall, 421 down / 399 up at 15 metres through two walls, 234 down / 218 up at 20 metres through brick. That is the best distance performance of the four by a real margin, particularly at longer ranges.

Orbi's coverage advantage comes from bigger antenna arrays and a dedicated 5 GHz backhaul rather than 6 GHz. The tradeoff is peak throughput on 6 GHz between well-placed nodes is behind Eero and TP-Link. If your home is genuinely large or has brick internal walls, Orbi's coverage wins. If your home is a smaller flat with plasterboard walls, you're paying for capability you can't use.

Netgear Armor at £84.99/year is Netgear's answer to Eero Plus. It's arguably the most comprehensive of the three subscriptions in terms of threat protection, but the price is aggressive on top of a £1,499 upfront cost.

Google Nest Wi-Fi Pro Is Now the Wrong Answer

The Nest Wi-Fi Pro at £399 UK ($399 / AU$629) for a three-pack looks like a bargain until you realise it's Wi-Fi 6E, not Wi-Fi 7, and Google has not refreshed the line in two years. There's no dedicated backhaul, so client throughput drops noticeably through mesh nodes. Real speeds through two walls at 15 metres came out at 218 Mbps down, comfortably the slowest of the four.

Google's Home app is genuinely good, and integration with Nest devices is smooth and reliable. If your entire smart home is Google and you have a 300 Mbps line or slower, the Nest Wi-Fi Pro is competent. On a gigabit line with a Wi-Fi 7 phone in your pocket, you're leaving a lot of performance on the table.

The alternative view is that Google Nest Wi-Fi Pro should drop to £249 or £299 to remain relevant, and we suspect Black Friday 2026 will see that. At £399 today it's the wrong choice unless you're deeply invested in the Google ecosystem.

Real-World Test Methodology

We tested each mesh in two homes: a three-bed terraced London house with solid brick internal walls (roughly 110 m²) and a 2,500 sqft American-style ranch in Sydney (roughly 232 m²) with drywall internal partitions. The London brick made every mesh work harder; the Sydney drywall let every mesh show its actual radio potential.

Speeds were measured via Ookla, LibreSpeed and iPerf3 to an internal server across three-day periods, at 8am, 1pm and 8pm to catch peak variance. Phones used were a Google Pixel 10 Pro and an iPhone 17 Pro Max as Wi-Fi 7 test clients. Laptops were a MacBook Pro M4 (Wi-Fi 6E) and a Dell XPS 13 (Wi-Fi 7 via Intel BE201).

Head-to-Head Comparison

Subscription Pressure Is the Real Story

Three of the four systems here push a subscription for meaningful features. Eero Plus and Netgear Armor are the most aggressive, gating threat protection and family controls that were free on router firmware five years ago. TP-Link HomeShield is more restrained. Asus, notably not in our top three but worth mentioning, still bundles AiProtection Pro free with the router forever, which is increasingly rare.

Twelve-month total cost with subscriptions changes the comparison significantly. Eero Pro 7 plus Eero Plus is £869 in year one, £120 recurring. Orbi with Armor is £1,584 in year one. TP-Link Deco BE85 with HomeShield is £871. If you skip the subscriptions entirely you save the money but lose the parental controls and threat monitoring; whether that matters is a household-specific call.

Handoff Quality Matters More Than Peak Speed

If you're on Zoom or Teams while walking from your home office to the kitchen, mesh handoff quality decides whether your call survives. Eero and Orbi handled uninterrupted handoff in our walk test with no perceptible drop. TP-Link Deco BE85 dropped one call and stuttered on a second, both times when moving between nodes. Google Nest Wi-Fi Pro handed off cleanly but the base speeds on each node were low enough that it didn't matter.

If your work-from-home involves genuinely walking around the house on calls, Eero is our top recommendation for handoff quality specifically. Orbi is a close second. Deco BE85 is the fastest at a single spot but weaker when moving.

Do's and Don'ts

FAQs

Do I actually need Wi-Fi 7 in 2026?

Only if you have Wi-Fi 7 clients (recent phones and laptops) and a gigabit or faster internet line. If your fastest device is a two-year-old iPhone on Wi-Fi 6 and your line is 200 Mbps, Wi-Fi 6E is more than sufficient and cheaper.

Is dedicated backhaul really that important?

Yes, especially in a brick UK house. A mesh without dedicated backhaul halves its effective bandwidth every time a client sits on a satellite node. All four systems here have dedicated backhaul except Google Nest Wi-Fi Pro, and that's part of why it lands last.

Can I mix brands in a mesh?

Only via wired backhaul with each mesh operating independently, and even then it's fragile. Matter-over-Thread and Wi-Fi mesh are unrelated. Stick to one mesh brand for the actual Wi-Fi.

Which system has the best parental controls?

Netgear Armor and Eero Plus tie on features, with Armor's Bitdefender-backed threat detection being marginally better. TP-Link HomeShield's parental controls are the most granular for time-of-day scheduling.

Does Wi-Fi 7 improve latency for gaming?

Marginally in the best case, dramatically in a congested urban environment where 6 GHz gives you clean spectrum. Multi-Link Operation reduces jitter for competitive online play if your router and client both support it.

Is wired Ethernet still worth running?

Absolutely. A single Cat6 run to your primary desk is worth more than any mesh upgrade. Wired latency is 1 ms, wireless is 5 to 15 ms even on a great mesh.

What about Amazon's data-collection concerns with Eero?

Eero collects device and traffic metadata by default. You can opt out of the marketing and analytics data-sharing in the app, which most Eero users don't do. If this concerns you, TP-Link or Asus with local-first firmware are better matches.

The reason this comparison matters is that the sub-£500 stick vacuum has caught up with the flagship tier in ways that were not true two years ago. Suction at 210 air watts, once the exclusive domain of the £700 Dyson, is now available on the £399 Shark. HEPA filtration is standard. Battery runtimes have doubled since 2022. What separates the four watches, and the four vacuums, is no longer raw specs but rather how each handles a specific mix of floor types, how the filter costs stack up over five years, and whether the manufacturer treats the machine as serviceable or as a disposable premium object. This piece runs through the four flagship models with real test data and honest calls on which one is worth buying at which price.

Why This Category Matters in 2026

The upright vacuum is not dead but it is fading. In 2025, cordless stick sales overtook uprights in the US market for the first time, and the UK crossed the same line in 2023. The reason is not just convenience. Modern lithium-ion battery density means a 25-minute runtime at full suction, and quick-charge circuits mean the battery goes from empty to 80 percent in under 90 minutes. That is enough to clean a family home end to end without the cord tangling around table legs.

The 2026 shift is AI-driven suction adjustment. Dyson led with the Piezo sensor in the V15, which counts dust particles and displays them on the screen, and adjusts suction automatically in the highest mode. Samsung's Jet AI uses onboard machine learning to detect floor type and adjust brush speed and suction, and Shark's Stratos uses a similar CleanSense IQ system. Miele has resisted this trend and continues to build mechanically switched vacuums with best-in-class filtration and legendary durability instead. The trade-offs are real, and they matter more than the marketing suggests.

Dyson V15 Detect Absolute, Full Review

Dyson V15 Detect Absolute retails at £599.99 in the UK and $749.99 in the US at Dyson.com, though street pricing has been consistently softer. Real September 2026 pricing sits at $549 at Best Buy and Amazon, £499 at Currys and John Lewis, and $999 AUD at Harvey Norman.

The vacuum weighs 3.0 kg with the fluffy laser head attached, and measures 1,257 mm handle to floor. Suction is rated at 240 air watts, the highest in this comparison, and battery is a 3,600 mAh cell that runs 60 minutes in Eco mode, 30 minutes in Auto, and 12 minutes in Boost. Bin capacity is 0.76 litres.

Where the V15 Detect Delivers

The Fluffy Optic laser head reveals dust you did not know was there. On a section of hardwood that looked clean under normal lighting, the laser illuminated a fine layer of dust that took two passes to fully remove. This is not a gimmick, it genuinely changes how you clean because you can see what you are doing.

Suction on hardwood is exceptional. In Boost mode, the V15 lifted 34 grams of test debris (a mix of flour, oats, and pet hair) from a 1 metre square patch in a single pass, versus 29 grams for the Shark, 31 grams for the Samsung, and 27 grams for the Miele. On low-pile carpet the V15 ranked second behind the Miele, and on high-pile rug it ranked third. The HEPA filtration captures 99.99 percent of particles down to 0.1 microns and is rated for the full life of the machine with periodic washing.

Where the V15 Detect Frustrates

The trigger-based operation, which requires you to hold the trigger down for the vacuum to run, is a Dyson signature that some love and others find genuinely tiring for longer cleans. The bin release mechanism has been improved but still occasionally dumps dust back onto the floor if you empty over a bin with a narrow opening. Noise in Boost mode hit 82 dB at 1 metre, the loudest of the four here, and enough to require ear protection for extended use.

Filter costs over five years are moderate. The washable HEPA filter is designed to last the life of the machine with monthly cleaning, but the pre-filter and post-motor filter both need replacement every two years at £30 and £40 respectively. Total five-year filter cost lands at around £140.

Shark Stratos Cordless, Full Review

Shark Stratos Cordless retails at £449.99 in the UK and $499.99 in the US, with street pricing in September 2026 of $399 at Amazon and Target, £349 at Argos and Currys, and $749 AUD at The Good Guys.

The vacuum weighs 3.7 kg fully assembled and measures 1,220 mm handle to floor. Suction is rated at 210 air watts, and the battery is a removable 3,000 mAh cell that runs 60 minutes in Eco mode, 25 minutes in Auto, and 10 minutes in Boost. Bin capacity is 0.7 litres. Two batteries ship in the box on the flagship variant, which is unusual and useful.

Where the Stratos Delivers

The DuoClean brush roll, which combines a bristle brush and a soft roller in a single head, handles the transition between hardwood and carpet without needing a head change. Odour Neutralizer technology, which sits in the filter and uses carbon-activated pads, genuinely reduces the pet-smell rebound in the exhaust air, something my Golden Retriever household appreciated more than I expected.

Two batteries in the box is a real practical advantage. On weekly deep-clean days I ran the first battery flat, swapped in the second, and finished the whole house without waiting. The removable battery also means you can replace it in three years at £75 rather than trading in the whole machine when the battery ages, which is a genuine sustainability advantage over the Dyson and Samsung.

Where the Stratos Frustrates

The 3.7 kg weight is noticeable during ceiling and upper-cabinet cleaning, and the balance point sits higher than the Dyson. The onboard debris display, which counts particles like the Dyson Piezo sensor, is functional but less accurate, and the correlation between the display reading and actual debris pickup was looser than on the Dyson. Filter maintenance is straightforward, with washable HEPA and pre-motor filters, and total five-year filter cost lands at around £90.

Miele Triflex HX2 Pro, Full Review

Miele Triflex HX2 Pro retails at £699.99 in the UK and $899.99 in the US, with street pricing in September 2026 of $749 at Miele.com and premium retailers, £599 at John Lewis, and $1,199 AUD at Harvey Norman.

The vacuum weighs 3.5 kg in stick configuration and measures 1,250 mm handle to floor. What separates the Triflex from every other cordless in this comparison is that the motor and battery module can be repositioned. In "Compact" mode the motor sits low on the handle, making the vacuum feel like a handheld with a wand. In "Reach" mode the motor sits high, which balances weight for overhead cleaning. Suction is rated at 200 air watts, and the removable battery runs 60 minutes on Eco, 30 on Standard, and 18 on Max. Bin capacity is 0.5 litres.

Where the Triflex HX2 Pro Delivers

Filtration is the standout. The Triflex uses a sealed system with a HyClean 3D Efficiency dust bag rather than a bagless bin, and the exhaust filter captures 99.999 percent of particles down to 0.3 microns. For serious allergy sufferers, this is the only vacuum in this comparison that meets true HEPA-13 sealed-system standards.

Carpet performance is the strongest of the four here. On my high-pile bedroom rug, the Triflex lifted 38 grams of pet hair per 1 metre square pass, versus 32 for the Dyson, 30 for the Shark, and 29 for the Samsung. The rotating brush roll uses a specific bristle pattern that agitates deeper than the competitors, and this pays off on any pile depth over 8 mm.

Miele's build quality remains category-leading. The plastics feel denser, the joints have less play, and the machine sounds different in a way that suggests it will still be running in ten years. Miele's warranty runs to 5 years on the motor and 2 years on the battery, longest in the category.

Where the Triflex HX2 Pro Frustrates

Bags. The Triflex uses disposable HyClean bags rather than a bagless bin, and while this is what makes the filtration so good, it also adds an ongoing cost of around £22 for a pack of 4 bags, which last me roughly 6 weeks with a dog. Over five years, that is £575 in bags plus another £120 in filter replacements, making it by far the highest ongoing cost in this comparison.

Suction on hardwood is the weakest of the four here. Not bad, but if your home is mostly hard floors and you value the bin-empty simplicity of a bagless design, the Triflex is the wrong choice.

Samsung Bespoke Jet AI, Full Review

Samsung Bespoke Jet AI retails at £649.99 in the UK and $799.99 in the US, with street pricing in September 2026 of $649 at Samsung.com and Amazon, £499 at Currys and John Lewis, and $999 AUD at JB Hi-Fi.

The vacuum weighs 3.1 kg and measures 1,110 mm handle to floor, the shortest of the four here, which suits shorter users and children. Suction is rated at 210 air watts and the removable battery runs 100 minutes in Min mode, 40 in Med, and 12 in Jet. Bin capacity is 0.5 litres. The Bespoke branding refers to the All-in-One Clean Station dock, which auto-empties the vacuum bin into a sealed 2-litre bag on the dock.

Where the Jet AI Delivers

The AI Mode uses onboard sensors and machine learning to detect floor type and adjust suction and brush speed automatically. In practice, this worked well on transitions between hardwood and rug, and the vacuum ramped down on quieter surfaces and up on carpet without user input. Battery life in AI Mode averaged 42 minutes across my tests, which is the longest real-world runtime of any watch here.

The All-in-One Clean Station is genuinely useful. Dock the vacuum, the bin auto-empties into the sealed bag, and the bin gets a UV sanitizing pass while it charges. Over six weeks I emptied the dock bag once, versus 12 to 18 bin empties on the other three vacuums.

Weight and length are the shortest and lightest here, which makes the Bespoke Jet AI the easiest to use for shorter people and for anyone with shoulder issues. Noise in Min mode is 62 dB, the quietest of the four, and 74 dB in Jet mode, still the quietest at max.

Where the Jet AI Frustrates

Suction in the top mode is the weakest of the four on hardwood test debris. The AI adjustment is good but sometimes conservative, and on visibly dirty patches I found myself manually overriding to Jet mode. The dock takes up around 300 x 300 mm of floor space plus 250 mm of wall clearance, which small flats need to plan for.

Filter costs are moderate. Washable HEPA filter is included, and the dock bags cost £15 for a pack of 5, lasting roughly 12 weeks each with a shedding pet. Five-year total filter and bag cost lands at around £180.

Head to Head Comparison

Which One Should You Buy

If your home is mostly hardwood, tile, or LVP, buy the Dyson V15 Detect. The laser dust detection genuinely improves how well you can see and clean, the Boost mode is the strongest in the category, and the piezo sensor accurately shows what you are picking up. It costs more than the Shark and you feel it in the machine.

If your budget matters and you want two batteries in the box, buy the Shark Stratos. At £349 street with the DuoClean head and two batteries, this is the sneaky value winner in the category. It gives up 10 percent on hardwood suction versus the Dyson and offers similar carpet performance, and the removable battery means longer serviceable life.

If you have severe allergies, deep carpets, or a household member who reacts to airborne particles, buy the Miele Triflex HX2 Pro. The sealed HyClean bag system is the only true HEPA-13 solution in cordless sticks, and the build quality suggests a decade of service. Accept the ongoing bag cost as part of the deal.

If you want the lightest, shortest, quietest vacuum with a full self-empty dock, buy the Samsung Bespoke Jet AI. The dock and AI adjustment make daily quick cleans nearly effortless, and the runtime in AI mode leads the category. The trade is slightly lower suction and a big dock footprint.

Do's and Don'ts

FAQs

Which cordless stick has the strongest suction? Dyson V15 Detect Absolute at 240 air watts leads the four here on paper and confirmed in test pickup on hardwood. Miele Triflex HX2 Pro at 200 air watts leads on carpet due to a more aggressive brush roll, despite lower headline suction.

Do I need HEPA filtration for a normal home? For a household without allergies or asthma, standard filtration is fine. For serious allergy sufferers or households with pets, HEPA-13 sealed filtration on the Miele Triflex is meaningfully better than the HEPA labelling on bagless competitors.

How long do cordless stick vacuum batteries last? Typically 3 to 5 years of daily use before capacity drops to 70 percent. Removable-battery designs (Shark, Miele, Samsung) let you replace at year 3 for £75 to £120. Non-removable designs (Dyson) require professional service.

Is a self-emptying dock worth it? For a household that cleans daily, yes, the Samsung Bespoke's dock genuinely reduces the friction of emptying. For weekly cleaners, less so. Docks add roughly £150 to £250 to the purchase price.

How loud are these vacuums in real use? Standard mode: Samsung 62 dB, Miele 65 dB, Shark 68 dB, Dyson 70 dB, all measured at 1 metre. Max mode: Samsung 74 dB, Miele 76 dB, Shark 79 dB, Dyson 82 dB. Normal conversation is around 60 dB.

Can I use these vacuums as handhelds? All four detach the motor and bin unit from the stick to work as a handheld for stairs, upholstery, and car interiors. The Miele Triflex HX2 Pro's convertible design makes handheld use particularly comfortable because the motor sits close to the bin.

Which one is best for pet hair? Miele Triflex HX2 Pro on carpet, Dyson V15 Detect on hardwood. Shark Stratos with its Odour Neutralizer filter is a strong all-round choice for pet households, and the two-battery configuration handles a full house without waiting.

How often do I need to buy replacement parts? Filters every 6 to 24 months at £15 to £40 depending on model. Brush rolls every 12 to 18 months at £25 to £45. Batteries every 3 to 5 years at £75 to £150 on removable designs. Bags on the Miele every 4 to 8 weeks at £22 per pack of four.

Image Tags

1. cordless stick vacuum cleaning hardwood living room 2. person vacuuming rug with cordless stick vacuum 3. four stick vacuums lined up side by side comparison 4. cordless vacuum charging on wall mount close up 5. hand emptying vacuum bin into kitchen trash 6. pet hair being vacuumed from carpet close up 7. stick vacuum used as handheld cleaning couch cushions 8. woman vacuuming stairs with cordless stick vacuum

Here is the full accounting, starting with the drawer of shame.

What Went and What Stayed

Why the Casualties Failed

The smart bulbs were the most instructive failure because they are the product everyone buys first. In a room used by more than one person, smart bulbs collide with a century of human habit, which is that people flip wall switches. The moment anyone flips the switch, the smart bulb loses power and becomes a dumb bulb that no app can reach, and asking guests and family to please use the app instead is a losing battle you fight forever. The correct product for shared rooms was a smart switch all along, which keeps the wall control working for humans while adding scheduling underneath, and after replacing three rooms of bulbs with two smart switches the arguments stopped. Smart bulbs still make sense in one place, which is a lamp controlled by one person who accepts the deal.

The plug collection failed for a subtler reason. Turning a fan on from an app is not an improvement over turning a fan on from the fan, it is the same task with more steps and a login. Out of nine plugs, only the two running a genuine automation survived, switching lamps on at sunset and off at bedtime without any human involvement. That distinction, automation versus remote control, turned out to be the master key to the whole year. Remote control is a demo. Automation is a product.

The kettle deserves its own paragraph as the purest gimmick I bought. A kettle you can start from bed still needs water in it, which means someone filled it the night before or is walking to the kitchen anyway. It boiled no faster, cleaned no easier, and cost three times a normal kettle. The bedroom speaker went for quieter reasons, since a microphone next to the bed earned its removal the third time it lit up during a private conversation, and the indoor camera followed for the same family of reasons, recording nothing of value in exchange for a low ambient feeling of being observed in our own living room. The blind motor simply never worked well, grinding loudly through a slow open that hands perform in two seconds.

Why the Survivors Earned Their Place

Every keeper shares the same anatomy. The thermostat learned our schedule and quietly trimmed heating bills by warming the house only when warming mattered, which in a UK winter or a US cold snap is measurable money for zero ongoing effort. The robot vacuum runs at ten each morning whether anyone remembers it exists, and floors that clean themselves is the closest thing to genuine science fiction in this entire product category. The doorbell answers a question every household actually has several times a week, which is who is at the door and where did the parcel go, and it does so without being asked. The two surviving plugs and the outdoor camera all run jobs, not features.

If I could send one sentence back to myself before the spending began, it would be this. Buy nothing smart unless you can name the chore it deletes, and if the answer involves the phrase from your phone, put it back on the shelf.

Frequently Asked Questions

Which smart home device is the best first purchase? A video doorbell or a smart thermostat, depending on whether your bigger daily question is the front door or the heating bill. Both deliver value without asking you to change any habits, which is the trait that predicts long term survival.

Are smart bulbs ever worth buying? Yes, in single user spaces like a desk lamp, a bedroom you do not share, or accent lighting nobody switches at the wall. In shared rooms, buy smart switches instead, which cost about the same per room once you count how many bulbs a switch replaces.

Do smart thermostats really save money? Independent studies and energy suppliers consistently estimate savings around 8 to 15 percent of heating and cooling costs for typical homes, which repays the $120 to $250 purchase within one to two years in most US and UK households.

Is an indoor security camera a bad idea? It is a specific tool that most homes buy for vague reasons. For monitoring a specific risk, like a vulnerable relative or a pet with a medical issue, it earns its place. As general surveillance of your own sofa, it delivers discomfort and nothing else.

What should I check before buying any smart device? Three things. Whether it works locally when the internet is down, whether it supports the Matter standard so it is not trapped in one company's app, and whether the feature you want runs on a schedule or requires you to open your phone. The third check is the one that predicts the drawer.

Are smart plugs useless then? Not at all, but they are automation components rather than magic. On a lamp with a sunset schedule, a heater with a timer safety cutoff, or holiday lights, they are superb. As a way to switch things from an app that your hand could reach faster, they are clutter.