Feed aggregator

Is Marvel Tokon: Fighting Souls one of the best fighting games of recent times? Yes, I would say so. Given its developer is Arc System Works — creators of Guilty Gear, Blazblue, and the excellent Granblue Fantasy Versus: Rising — it was a pretty safe bet that we'd be served a high quality fighting game with mountainous respect and reverence for Marvel and its roster of supers, mutants, and weirdos.

Review info

Platform reviewed: PC
Available on: PS5, PC
Release date: August 6, 2026

I love fighting games, but I'm not usually one for tag fighters — often because I find learning more than one character intimidating and time-consuming. So, when Marvel Tokon: Fighting Souls was first revealed last year as a 4v4 tag fighter, I'd already resigned myself to thinking: 'well, that's probably not for me.'

Not two, not three, but four? You're having a laugh.

I needn't have worried so much. Arc System Works has designed Fighting Souls such that learning a quartet of fighters inside and out is absolutely not necessary. It's possible to play that way at a higher level, of course. But as your team only has a single health bar, you can rock one character as your main, while relying on the remaining three as assist moves to help you out in a pinch.

It helps, too, that Marvel Tokon: Fighting Souls has a fantastic onboarding process for new players. The game's story mode is broken into five episodes — each focused on a group of heroes (the Avengers, X-Men, etc.) or the villainous Knights of Doom. Episode chapters are structured to introduce you to fighters and their strengths, or otherwise various important system mechanics. Training tools are similarly great, with a fully-featured training mode, combo trials, and an ingenious 'level-up missions' category that tasks you with creating your own combos to reach a damage threshold.

All good so far, but I unfortunately have to stress that Tokon's immediate future looks perilous. The PC version of the game is currently facing myriad issues, many of which seem to be a direct result of corporate interference.

At the time of writing, Arc System Works is hard at work deploying patches to improve PC matchmaking and performance, but I'm not sure there's a fix for intrusive anti-cheat bloatware and forcing PlayStation Network account creation to access its online features.

Absolutely champion

(Image credit: Sony / Marvel)

Enough doom and gloom for now. I'd rather chat about all the elements that I feel make Marvel Tokon: Fighting Souls such a wonderful game for a bit. For me, it started with its single-player story mode.

Titled 'Episodes', Tokon's story is broken into five distinct, well, episodes. Each is centered around one of the game's five main teams. The Fighting Avengers, for example, are your stalwart heroes like Captain America and Black Panther.

The Samurai Outriders are made up of anti-heroes and tricksters like Deadpool and Loki. Meanwhile, the Knights of Doom bring a more villainous flavor, stacked with evildoers like Doctor Doom, Green Goblin, and Carnage.

Best bit

(Image credit: Sony / Marvel)

I'm in love with Marvel Tokon: Fighting Souls' eclectic moment-to-moment tag bouts, but a special shoutout has to go to the game's wicked cel-shaded presentation. Marvel icons have been given a distinctly Arc System Works flavor — their appearance, animation, and super move cut-ins are all top-tier stuff.

The premise is simple. An intergalactic fighter called the Champion is holding Earth hostage, and the aforementioned five teams have been selected to participate in a tournament to decide the planet's future. Each episode culminates in a fight with the Champion, and individual chapters are bookended with stunning comic book cutscenes, with each episode even featuring a unique artist and colorist.

This lends all of them a unique sense of identity, and helps to keep things feeling fresh as you progress. Each episode is about an hour long, and all are worth playing through as you'll not only get a feel for each individual character, you will also unlock colors for a bit of added flair for your faves. Oh, and as a bonus, you'll also unlock the Champion as a playable character upon clearing your first episode.

There are a lot of fun character moments, too. Ms. Marvel is often adorably starstruck by the heroes she encounters. The X-Men are flabberghasted that Magneto would team up with an egotistical maniac like Doctor Doom. There's a cute running gag between episodes where everyone gives Loki the side eye, despite a seeming earnestness from the God of Lies to genuinely turn over a new leaf in the search for his brother Thor (who is currently mysteriously absent from the game's playable roster).

Overall, Episodes are a great way to kick off your Tokon journey. It'll help you find characters you enjoy playing, and present a charming story that's well worth playing through just to experience its wonderful artwork and voiced performances.

Assist me

(Image credit: Sony / Marvel)

Marvel Tokon: Fighting Souls is remarkably easy to just pick up and play, and I reckon you'll have a blast in local play with friends for casual couch bouts. Movement, special moves, assists, and supers are all straightforward to execute. For specials in particular, you have options for simple or motion inputs with the d-pad. Motion-inputted special moves (such as quarter-circle-forward) do around 10% more damage on average, so you'll have a bit of an edge if you can master weaving them into your gameplan.

The 4v4 setup can be wonderfully chaotic; there's no limit to the number of characters that can appear on-screen, so it's very possible to spam the assist button to throw the kitchen sink at your opponent in a pinch.

Characters you place in your assist slots are highly versatile, and can be individually called upon by pressing the assist button and the corresponding direction on the d-pad. One assist slot is for projectile, for example, while another is a vertical for anti-airs. This is universal across the cast, so you don't have to be precious about which characters you choose if you're only planning on maining one or two.

Elsewhere, you can perform an assist counter while guarding to get your opponent off of you, or even swap to one of your assists to directly control them. There's a ton you can do with assists, and learning to incorporate them in your playstyle is a huge part of the learning process. It sounds like a lot, and it is, and there's much more assists can do, like extending your combos and taking part in a massively damaging last round super. It might take a bit to wrap your head around, but thankfully, the tools are there to help you get to grips with assists and more besides.

Trained to perfection

(Image credit: Sony / Marvel)

Marvel Tokon: Fighting Souls has some of the best tutorialization I've ever seen in a fighting game. You of course have a fully fleshed out training mode, where you can freely practice combos and setups against a dummy. There is also a comprehensive tutorial section known as the Battle Guide, that'll walk you through basics like movement all the way to advanced techniques like wakeup options and blocking cross-ups.

Of particular note is Character Lessons, which includes primers on every character playable in the game. These start off with basic move introductions, as well as the properties of each characters' unique button. For example, Storm's Tempest unique lets her control the wind to push herself and her opponent around. Magik's unique has her placing a portal that she can instantly warp to and perform attack out of to surprise the opponent.

Lessons then offer a selection of combos for you to try out for each character, many weaving in assists, special moves, and supers. Finally, the highlight of Character Lessons for me are the Level-Up Missions, which task you with performing a combo to meet a certain amount of damage. These don't hold your hand much, with little guidance outside of telling you which button to start your combo with, or to include a special move at some point. They're great fun to experiment with, and even unlock colors for your characters, too.

Once you feel sufficiently warmed up, I highly recommend jumping into Arena mode for some offline quickplay. This is essentially the game's Arcade mode, in which you play through a series of five matches. You'll get to choose your route as you go through, unlocking some lovely concept art for your efforts.

There's also Survival mode; a 100-match gauntlet with some light roguelite elements. Between matches in Survival mode, you'll get to choose and upgrade Support Souls — cards that offer various benefits like increased movement speed or attack power. Getting through all 100 is no easy task, though, with several significant difficulty spikes on your way to the top. But again, your efforts will be rewarded with unique character colors if you make it to the end.

True believers

(Image credit: Sony / Marvel)

If you're familiar with Arc System Works' back catalog, it should come as no surprise that Marvel Tokon: Fighting Souls is a mind-bendingly gorgeous fighting game.

The developer is the master of modern cel-shading, and the roster of Marvel icons look even sharper here than characters do in Guilty Gear Strive and Dragonball FighterZ. Not that those games are slouches at all in the visual department, but Tokon by comparison looks impressively pristine at all times.

There aren't a ton of characters or stages at launch (I personally think 21 characters is a bit slim given the 4v4 setup), but when they look and animate as well as they do, I can forgive the base roster being as svelte as it is by tag fighter standards.

Every character is brimming with personality and charming little touches. Doom's cartoonish aptitude for evil is perfectly captured in his intimidating slow walk and thousand-yard stare. Ms. Marvel's attacks feel peppy, snappy, and excitable. And Deadpool? Well, almost every move in his arsenal is a loving reference to other iconic fighting games, and there are some impressively deep cuts therein.

The game's background stages are a real draw, too. The New York City stage is teeming with detail and references to the wider world of Marvel. The X-Mansion tends to populate the further a match progresses, with iconic X-Men like Psylocke, Colossus, and Gambit appearing in the background. If you're a Marvel head, you'll likely love the amount of respect and admiration Arc System Works has for the brand.

Pride before a fall

(Image credit: Sony / Marvel)

Marvel Tokon: Fighting Souls is another slam-dunk for Arc System Works, then. With it and the recent Marvel vs. Capcom Fighting Collection: Arcade Classics compilation, I'm ecstatic to see Marvel active again in the realm of fighting games. But every story has its villain, and for Tokon, it's not the Champion. It's Sony.

I hate to admit that I'm concerned for Marvel Tokon: Fighting Souls' future. On PC, the game's anti-cheat DRM worsens performance. Furthermore, Linux and mod support are non-starters, and because online play requires a PlayStation Network account, countries where the service isn't supported (there are well over 100 of them) can't access network features at all. Throw in lopsided netcode that discourages crossplay between PS5 and PC, and you have a recipe for an exceptionally messy launch month largely thanks to publisher meddling.

For these reasons, I can't give Marvel Tokon: Fighting Souls a glowing recommendation, as much as I would like to. At least if you were planning to pick up the game on PC. I want Tokon to have a long, bright future, but I worry that too much damage has already been done. Now, the state of the game has improved since launch in regards to performance and matchmaking quality, thanks to a small handful of fixes deployed by Arc System Works.

I just can't help but feel anxious about this rocky PC launch. Especially with the recent news that Riot Games is shuttering development on its League of Legends fighting game, 2XKO. It shows that huge companies absolutely aren't afraid to completely tear down a good thing if it isn't generating satisfactory revenue or player count. If folks walk away from Marvel Tokon: Fighting Souls because of the issues it's facing on PC, then what's stopping Sony from doing the same?

I hope I'm just overthinking it. Marvel Tokon: Fighting Souls is one of the most fun and unique fighting games I've played in some time, and I hope its candle won't be snuffed out before it truly gets the chance to shine its brightest.

Should I play Marvel Tokon: Fighting Souls?Play it if...

You love the Marvel tag fighters of old
Living up to Marvel vs. Capcom's legacy is no easy task. And while Marvel Tokon: Fighting Souls is doing its own thing with its own roster, it's nonetheless continued proof that Marvel and tag fighters are a winning combination.

You're completely new to fighting games
Learning fighting games can be daunting, and Tokon certainly has a high skill ceiling with its tagging and various system mechanics. But elements like auto combos, easy special move inputs, and a recognizable roster of Marvel icons make this game incredibly newcomer-friendly.

Don't play it if...

You want the PC version
This hopefully won't be a dealbreaker for long especially with the impressive pace Arc System Works is deploying fixes. But PC performance and matchmaking issues are still ongoing, and it may be worth waiting a few months for the Steam version to catch up with its PS5 counterpart in overall quality.

Accessibility features

At launch, Marvel Tokon: Fighting Souls has a small handful of accessibility features, largely utilizing text-to-speech options.

You can enable text-to-speech and adjust its volume and speed. You can also choose to enable or disable subtitles that display during battle entrances and conclusions.

How I reviewed Marvel Tokon: Fighting Souls

I played Marvel Tokon: Fighting Souls on PC for approximately 15 hours for this review. I completed the episodic story mode in its entirety, cleared most Arcade mode routes, and put a good chunk of hours into training mode, character lesson guides, and casual online play.

I tested the game with the GameSir G7 Pro controller, as well as the leverless Razer Kitsune. I also hooked up my RIG 900 Max HS gaming headset to experience the game's excellent audio and wonderfully upbeat soundtrack.

First reviewed August 2026

The world largely moved on from the Rohingya crisis. But the exodus has not stopped, aid is shrinking and a generation of refugees is growing up with little prospect of going home.

(Image credit: Valeria Mongelli)

On its own, an air fryer is a low-effort, efficient kitchen appliance. But add the right accessories and you can take things to another level, enabling your trusty fryer to tackle a wider range of cooking tasks, and deliver even more delicious results. Plus, if you're armed with the right kit, there'll be even less effort required from you.

As a former TechRadar Homes Editor — and proud owner of two air fryers — I have a good idea of the accessories worth investing in. Well, I say 'investing' but most of my picks are under a tenner, so it's more an investment in terms of cupboard space than cash. I've rounded up my top recommendations below, with sections for air fryer liners (wave goodbye to messy cleanup), utensils (including a £5 pick for extra-crispy results) and inserts (to increase capacity without buying a bigger air fryer).

If you don't already have an air fryer, head to our general best air fryer guide for our top recommendations.

Air fryer liners

Good silicone basket liners are an excellent investment for any air fryer-owner, mainly because they're much easier to clean (and especially if you have a dishwasher). I use these all the time — in fact, the only time I wouldn't use one of these would be if I was cooking something I want to get really crispy underneath. I avoid them for cheese on toast, for instance, to avoid a soggy bottom.

Choose from solid or collapsible liners — the latter is a win when it comes to storage, but the fold lines are a potential weak point, so be prepared for them to break more quickly than a solid option. To avoid disaster when trying to remove your hot food from its basket, seek out liners with sturdy handles.

You can also use basket liners to add versatility to what you can cook. For instance, there are brownie trays, cupcake cases and waffle trays in the mix here. I've also included a disposable liner option, for situations where washing up is going to be too tricky.

Tower Set of 2 Silicone Air Fryer Rectangular Solid Trays

Tower Set of 2 Silicone Air Fryer Square Foldable Trays

Tower Round Solid Silicone Tray with Divider

Lakeland 2x Air Fryer Silicone Brownie Moulds

Lakeland 2x Air Fryer Round Silicone Moulds

ProCook Silicone Air Fryer Liner Rectangle 19.5 X 12.5cm

Generic Air Fryer Liners for Dual Zone Air Fryers (100pcs)

Air fryer inserts

A genius trick to to increase your usable cooking space is by adding an insert that creates a second tier within your air fryer basket. If you have a relatively large basket in the first place, you can even stack multiple racks to cook several layers of food at once. Hot air should still circulate efficiently — and especially so if you use a rack with vent holes — to deliver crispier, more evenly cooked food than you'd get if you piled everything up inside.

Below are a few options. I've also included a mesh basket — this is designed to be used in a similar way to the liners above; to make cleaning easier. However, the more open structure should help deliver crispier results. Some even have divots into which skewers can rest, for 360-degree cooking. Before purchasing, it's vital to measure the inside of your cooking basket(s) and check the product dimensions carefully, to ensure the inserts will fit in there.

Tefal Tefal 4-Piece Air Fryer Accessory Set, Non-Stick Carbon Steel

Tala Stainless Steel 2 Tier Air Fryer Rack & 4 Skewers

Kitchen Master Rectangle Stainless Steel Air fryer Rack

Tower 8 piece Silicone Air Fryer Accessory Set

Lakeland Air Fryer Mesh Basket

Air fryer utensils

Finally, there are a number of utensils that'll make cooking with an air fryer simpler and less messy — not to mention lowering the chance of burning yourself or your kitchen surfaces. If you don't already have one, I'd suggest investing in a heatproof mat or trivet. It doesn't need to be anything special; it just gives you somewhere to put your fryer basket down when you're removing the hot food. I've had too many accidents when trying to do this one-handed, or with the basket balancing precariously on the hob top.

Silicone utensils including tongs, scoops and spatulas allow you to remove food efficiently without risking scratches on the non-stick surface. I've also included a pastry brush — I use one of these to lightly coat food in oil before cooking, so it goes extra-crispy, but without using too much oil.

Finally, there's a meat thermometer that'll help you ensure your cuts of meat are cooked through before serving — this is especially handy if you're still getting used to the different timings compared to a full-sized cooker. And because air fryers really shine when it comes to chips, I've thrown in a potato chipper so you can make your own.

OXO Good Grips Roast Scoop

ProCook Wood & Silicone Spoonula Blue 26cm

ProCook Silicone Pastry Brush Green 26cm

procook Silicone Tongs Green 29cm

Potato Chipper Stainless Steel

Tower Vortx 17-Piece Air Fryer Oven Baking Bundle

Lakeland Heat Resistant Air Fryer Mat

Folding Trivet — Sage

ProCook Silicone Trivet — Duck Egg Blue

ThermoMaven Digital Meat Thermometer

On its own, an air fryer is a low-effort, efficient kitchen appliance. But add the right accessories and you can take things to another level, enabling your trusty fryer to tackle a wider range of cooking tasks, and deliver even more delicious results. Plus, if you're armed with the right kit, there'll be even less effort required from you.

As a former TechRadar Homes Editor — and proud owner of two air fryers — I have a good idea of the accessories worth investing in. Well, I say 'investing' but most of my picks are under a tenner, so it's more an investment in terms of cupboard space than cash. I've rounded up my top recommendations below, with sections for air fryer liners (wave goodbye to messy cleanup), utensils (including a £5 pick for extra-crispy results) and inserts (to increase capacity without buying a bigger air fryer).

If you don't already have an air fryer, head to our general best air fryer guide for our top recommendations.

Air fryer liners

Good silicone basket liners are an excellent investment for any air fryer-owner, mainly because they're much easier to clean (and especially if you have a dishwasher). I use these all the time — in fact, the only time I wouldn't use one of these would be if I was cooking something I want to get really crispy underneath. I avoid them for cheese on toast, for instance, to avoid a soggy bottom.

Choose from solid or collapsible liners — the latter is a win when it comes to storage, but the fold lines are a potential weak point, so be prepared for them to break more quickly than a solid option. To avoid disaster when trying to remove your hot food from its basket, seek out liners with sturdy handles.

You can also use basket liners to add versatility to what you can cook. For instance, there are brownie trays, cupcake cases and waffle trays in the mix here. I've also included a disposable liner option, for situations where washing up is going to be too tricky.

Tower Set of 2 Silicone Air Fryer Rectangular Solid Trays

Tower Set of 2 Silicone Air Fryer Square Foldable Trays

Tower Round Solid Silicone Tray with Divider

Lakeland 2x Air Fryer Silicone Brownie Moulds

Lakeland 2x Air Fryer Round Silicone Moulds

ProCook Silicone Air Fryer Liner Rectangle 19.5 X 12.5cm

Generic Air Fryer Liners for Dual Zone Air Fryers (100pcs)

Air fryer inserts

A genius trick to to increase your usable cooking space is by adding an insert that creates a second tier within your air fryer basket. If you have a relatively large basket in the first place, you can even stack multiple racks to cook several layers of food at once. Hot air should still circulate efficiently — and especially so if you use a rack with vent holes — to deliver crispier, more evenly cooked food than you'd get if you piled everything up inside.

Below are a few options. I've also included a mesh basket — this is designed to be used in a similar way to the liners above; to make cleaning easier. However, the more open structure should help deliver crispier results. Some even have divots into which skewers can rest, for 360-degree cooking. Before purchasing, it's vital to measure the inside of your cooking basket(s) and check the product dimensions carefully, to ensure the inserts will fit in there.

Tefal Tefal 4-Piece Air Fryer Accessory Set, Non-Stick Carbon Steel

Tala Stainless Steel 2 Tier Air Fryer Rack & 4 Skewers

Kitchen Master Rectangle Stainless Steel Air fryer Rack

Tower 8 piece Silicone Air Fryer Accessory Set

Lakeland Air Fryer Mesh Basket

Air fryer utensils

Finally, there are a number of utensils that'll make cooking with an air fryer simpler and less messy — not to mention lowering the chance of burning yourself or your kitchen surfaces. If you don't already have one, I'd suggest investing in a heatproof mat or trivet. It doesn't need to be anything special; it just gives you somewhere to put your fryer basket down when you're removing the hot food. I've had too many accidents when trying to do this one-handed, or with the basket balancing precariously on the hob top.

Silicone utensils including tongs, scoops and spatulas allow you to remove food efficiently without risking scratches on the non-stick surface. I've also included a pastry brush — I use one of these to lightly coat food in oil before cooking, so it goes extra-crispy, but without using too much oil.

Finally, there's a meat thermometer that'll help you ensure your cuts of meat are cooked through before serving — this is especially handy if you're still getting used to the different timings compared to a full-sized cooker. And because air fryers really shine when it comes to chips, I've thrown in a potato chipper so you can make your own.

OXO Good Grips Roast Scoop

ProCook Wood & Silicone Spoonula Blue 26cm

ProCook Silicone Pastry Brush Green 26cm

procook Silicone Tongs Green 29cm

Potato Chipper Stainless Steel

Tower Vortx 17-Piece Air Fryer Oven Baking Bundle

Lakeland Heat Resistant Air Fryer Mat

Folding Trivet — Sage

ProCook Silicone Trivet — Duck Egg Blue

ThermoMaven Digital Meat Thermometer

With a bank holiday weekend almost here in the UK, Currys has launched another of its Epic Deals events, featuring savings of up to 40% on loads of top tech products. I've done my usual thing of searching through the sale and picked out 12 of the best deals I'd recommend on TVs, laptops, appliances, headphones, and more.

• View the full Currys Epic Deals sale

As the back-to-school sales are still ongoing, it's no surprise that laptops make up a large part of this sale. If there's one offer that stands out amongst everything, though, it's this Acer Aspire 14 for £549 (was £699).

It's a great price for a modern, powerful laptop that features one of the newer, more efficient Snapdragon X processors, plus 16GB of RAM and a 512GB SSD — excellent specs for college, university, or everyday use. As well as high levels of performance for work and productivity needs, it also boasts an amazing 28 hours of battery life, which is great if you're always on the move.

Other deals I like the look of in the Currys Epic Deals sale include this five-star LG C6 48-inch OLED TV for £1,199 (was £1,399), these great value Bose QuietComfort Earbuds for £119 (was £179), and the popular Dyson V8 Vacuum Cleaner for £239 (was £349).

You can check out all of my top picks below, or have a browse through all the major categories from the sale to see if anything else looks good to you. Expect these offers to be available for at least a week, so you've got a bit of time before you need to hit that buy button.

Currys bank holiday sale — quick links

• Appliances: up to 40% off Bosch & Beko
• Headphones: JBL & Bose from £39.99
• Health & Beauty: up to £150 off Philips & Braun
• Kitchen: up to £550 off Tefal & De'Longhi
• Laptops: from £179
• Outdoors: up to 40% off DIY & tools
• Phones: up to £100 off Apple & Samsung
• Tablets: Lenovo & Samsung from £159
• TVs: from £109.99
• Vacuums: up to £150 off Dyson & Vax

Currys Epic Deals sale — my top 12 picks

Aspire 14: was £699 now £549
Display - 14 inches
Processor - Snapdragon X
RAM - 16GB
Storage - 512GB

This is a great price for a relatively modern laptop that comes with a well-performing and efficient Snapdragon X CPU. On top of that, you get 16GB of RAM, a 512GB SSD, and incredible battery life of up to 28 hours, which represents good value for money overall. I like it as a light, portable and capable all-around everyday machine.View Deal

As some of the best noise-cancelling earbuds on the market, it's great to see this £60 price drop on the Bose QuietComfort Earbuds. For the money, you get amazing active noise cancellation and beautifully balanced sound, all topped off with a gorgeous design and impressive 8.5-hour battery life. Altogether, these are some of the best mid-range earbuds available now for a record-low price.View Deal

Aspire Go 15: was £649 now £449
Display - 15.6 inches
Processor - Intel Core i5
RAM - 16GB
Storage - 512GB

Here's a solid and affordable Windows laptop available at Currys. It's an older machine, but there's no denying it's a solid entry-level choice for the price, including an Intel Core i5 CPU, 16GB of RAM, and a 512GB SSD for storage. It's a great buy if you want a good laptop with decent battery life for everyday use, work, or school.View Deal

The Dyson V8 is now older, but it remains one of the top affordable vacuums and also one of the best Dyson vacuums – period. This still more-than-capable vacuum comes with a couple of essential attachments for cleaning around the home. Given its age and basic features, it's a struggle to recommend it at its full price, but now that it's discounted so heavily at Currys, it's an easy recommendation.View Deal

We thought that the iPhone 17 Pro Max was the best iPhone ever in our review, and here's a chance to save some cash on this latest model at Currys. With a fresh new design, an outstanding camera array, a powerful chipset, and excellent battery life, it's everything you'd want from a premium Apple phone.View Deal

Promising simplicity of use, the De'Longhi Magnifica S features an easy-to-use touch control panel that lets you brew two cups simultaneously with minimal effort. It also has a silent integrated grinder with 13 settings, plus a handy milk frother too, so it’s a good all-in-one solution at a very affordable price.View Deal

Bravia 8 II 55-inch OLED TV: was £1999 now £1899
The latest version of Sony's Bravia 8 is an incredible OLED TV that boasts some of the best picture quality to date, according to our review. The display was only released this year and this is the biggest discount yet. Sure, it's still quite the investment, but you do get a picture quality that's bright, vivid, and detailed, with strong in-built sound as well. It's perhaps not the best pick for gamers, though, given it lacks some features compared to the competition.View Deal

DeskJet 2910 All-in-One Wireless Inkjet Printer: was £54.99 now £39.99
If you need a cheap and cheerful printer, this DeskJet 2910 is a great budget choice that supports wireless printing, scanning, and copying. This bundle also includes three months of HP Instant Ink, so you can save cash on your first few rounds of prints.View Deal

C6 48-inch OLED TV: was £1399 now £1199
LG reclaimed the top spot on our best TV guide with the C6, as it boasts fantastic overall picture quality and excellent all-around performance, making it the number one pick for most people's needs. Contrast is top-tier, with vivid colors and rich blacks, while motion is handled expertly. No matter if you're watching films, streaming TV shows, playing games, or watching sports, it's up to the job.View Deal

Yoga Slim 7: was £999 now £699
Display: 14 inch OLED
Processor: Intel Core Ultra 5
RAM: 16GB
Storage: 512GB

If display quality is just as important to you as power, then strongly consider this Lenovo Yoga deal at Currys. It comes with a top-end OLED screen that offers a clear and sharp image – whether you're working on documents or streaming your favourite TV shows. Specs are solid elsewhere, too, including a capable Intel Core Ultra 5 CPU and 16GB of RAM that ensure good overall performance.View Deal

WV2 Plus Window Vac: was £69.99 now £49.99
You can save a decent £20 on the Karcher WV2 in the Currys bank holiday sale, bringing down the basic window vac to under £50. The rechargeable lithium-ion battery boasts a 35-minute run time, which equals about 105m² on a single charge. That’s more than enough to do a small home or flat multiple times over.View Deal

A clear party piece or ideal addition to your World Cup viewing party if there ever was one, this PerfectDraft Beer Dispenser connects up to compatible 6-litre kegs of your favourite tipple and pours them like you were just down the pub. The convenient LCD screen shows how much is left and when the beer or cider is at the right temperature, while its compact design and dishwasher-safe components mean it's easy to store and clean.View Deal

With a bank holiday weekend almost here in the UK, Currys has launched another of its Epic Deals events, featuring savings of up to 40% on loads of top tech products. I've done my usual thing of searching through the sale and picked out 12 of the best deals I'd recommend on TVs, laptops, appliances, headphones, and more.

• View the full Currys Epic Deals sale

As the back-to-school sales are still ongoing, it's no surprise that laptops make up a large part of this sale. If there's one offer that stands out amongst everything, though, it's this Acer Aspire 14 for £549 (was £699).

It's a great price for a modern, powerful laptop that features one of the newer, more efficient Snapdragon X processors, plus 16GB of RAM and a 512GB SSD — excellent specs for college, university, or everyday use. As well as high levels of performance for work and productivity needs, it also boasts an amazing 28 hours of battery life, which is great if you're always on the move.

Other deals I like the look of in the Currys Epic Deals sale include this five-star LG C6 48-inch OLED TV for £1,199 (was £1,399), these great value Bose QuietComfort Earbuds for £119 (was £179), and the popular Dyson V8 Vacuum Cleaner for £239 (was £349).

You can check out all of my top picks below, or have a browse through all the major categories from the sale to see if anything else looks good to you. Expect these offers to be available for at least a week, so you've got a bit of time before you need to hit that buy button.

Currys bank holiday sale — quick links

• Appliances: up to 40% off Bosch & Beko
• Headphones: JBL & Bose from £39.99
• Health & Beauty: up to £150 off Philips & Braun
• Kitchen: up to £550 off Tefal & De'Longhi
• Laptops: from £179
• Outdoors: up to 40% off DIY & tools
• Phones: up to £100 off Apple & Samsung
• Tablets: Lenovo & Samsung from £159
• TVs: from £109.99
• Vacuums: up to £150 off Dyson & Vax

Currys Epic Deals sale — my top 12 picks

Aspire 14: was £699 now £549
Display - 14 inches
Processor - Snapdragon X
RAM - 16GB
Storage - 512GB

This is a great price for a relatively modern laptop that comes with a well-performing and efficient Snapdragon X CPU. On top of that, you get 16GB of RAM, a 512GB SSD, and incredible battery life of up to 28 hours, which represents good value for money overall. I like it as a light, portable and capable all-around everyday machine.View Deal

As some of the best noise-cancelling earbuds on the market, it's great to see this £60 price drop on the Bose QuietComfort Earbuds. For the money, you get amazing active noise cancellation and beautifully balanced sound, all topped off with a gorgeous design and impressive 8.5-hour battery life. Altogether, these are some of the best mid-range earbuds available now for a record-low price.View Deal

Aspire Go 15: was £649 now £449
Display - 15.6 inches
Processor - Intel Core i5
RAM - 16GB
Storage - 512GB

Here's a solid and affordable Windows laptop available at Currys. It's an older machine, but there's no denying it's a solid entry-level choice for the price, including an Intel Core i5 CPU, 16GB of RAM, and a 512GB SSD for storage. It's a great buy if you want a good laptop with decent battery life for everyday use, work, or school.View Deal

The Dyson V8 is now older, but it remains one of the top affordable vacuums and also one of the best Dyson vacuums – period. This still more-than-capable vacuum comes with a couple of essential attachments for cleaning around the home. Given its age and basic features, it's a struggle to recommend it at its full price, but now that it's discounted so heavily at Currys, it's an easy recommendation.View Deal

We thought that the iPhone 17 Pro Max was the best iPhone ever in our review, and here's a chance to save some cash on this latest model at Currys. With a fresh new design, an outstanding camera array, a powerful chipset, and excellent battery life, it's everything you'd want from a premium Apple phone.View Deal

Promising simplicity of use, the De'Longhi Magnifica S features an easy-to-use touch control panel that lets you brew two cups simultaneously with minimal effort. It also has a silent integrated grinder with 13 settings, plus a handy milk frother too, so it’s a good all-in-one solution at a very affordable price.View Deal

Bravia 8 II 55-inch OLED TV: was £1999 now £1899
The latest version of Sony's Bravia 8 is an incredible OLED TV that boasts some of the best picture quality to date, according to our review. The display was only released this year and this is the biggest discount yet. Sure, it's still quite the investment, but you do get a picture quality that's bright, vivid, and detailed, with strong in-built sound as well. It's perhaps not the best pick for gamers, though, given it lacks some features compared to the competition.View Deal

DeskJet 2910 All-in-One Wireless Inkjet Printer: was £54.99 now £39.99
If you need a cheap and cheerful printer, this DeskJet 2910 is a great budget choice that supports wireless printing, scanning, and copying. This bundle also includes three months of HP Instant Ink, so you can save cash on your first few rounds of prints.View Deal

C6 48-inch OLED TV: was £1399 now £1199
LG reclaimed the top spot on our best TV guide with the C6, as it boasts fantastic overall picture quality and excellent all-around performance, making it the number one pick for most people's needs. Contrast is top-tier, with vivid colors and rich blacks, while motion is handled expertly. No matter if you're watching films, streaming TV shows, playing games, or watching sports, it's up to the job.View Deal

Yoga Slim 7: was £999 now £699
Display: 14 inch OLED
Processor: Intel Core Ultra 5
RAM: 16GB
Storage: 512GB

If display quality is just as important to you as power, then strongly consider this Lenovo Yoga deal at Currys. It comes with a top-end OLED screen that offers a clear and sharp image – whether you're working on documents or streaming your favourite TV shows. Specs are solid elsewhere, too, including a capable Intel Core Ultra 5 CPU and 16GB of RAM that ensure good overall performance.View Deal

WV2 Plus Window Vac: was £69.99 now £49.99
You can save a decent £20 on the Karcher WV2 in the Currys bank holiday sale, bringing down the basic window vac to under £50. The rechargeable lithium-ion battery boasts a 35-minute run time, which equals about 105m² on a single charge. That’s more than enough to do a small home or flat multiple times over.View Deal

A clear party piece or ideal addition to your World Cup viewing party if there ever was one, this PerfectDraft Beer Dispenser connects up to compatible 6-litre kegs of your favourite tipple and pours them like you were just down the pub. The convenient LCD screen shows how much is left and when the beer or cider is at the right temperature, while its compact design and dishwasher-safe components mean it's easy to store and clean.View Deal

Voters in at least six states will soon decide ballot measures related to voter ID — the largest number of voter ID measures in a single year. The tally includes three prominent swing states.

(Image credit: Chris Carlson)

Maybe money does not explain everything, but it explains a lot. Goldman Sachs estimates the AI build-out will require roughly $7.6 trillion in capital over the next five years across compute, data centers and power. It’s unlikely that subscriptions alone will cover the bill, which is why ads are coming to AI.

OpenAI started rolling out ads in some ChatGPT tiers this year and forecasts $2.5 billion in ad revenue in 2026 and $100 billion by 2030. At an ad tech event this summer, OpenAI's chief revenue officer Denise Dresser dropped the hedging and said the company is "clearly in the advertising business now."

Perplexity, on the other hand, shut down its ad business entirely over user trust. Similar products, same opportunity, opposite conclusions.

This conflict is not new and the open web fought this battle 15 years ago, when ad blocking became mainstream.

Now, the AI industry is on track to repeat every mistake of the so-called ad blocking wars, unless it studies them instead.

Lesson one: users revolt faster than big platforms think

When ad formats became more intrusive - because again, money! - the industry assumed users would just accept a bad experience. Instead, users fought back and installed software to get rid of aggressive online ads. In what technology commentator Doc Searls called “the biggest boycott in world history”, ad blocking went from being a niche tool to a mainstream product.

And we see the same dynamic playing out again. Within weeks of sponsored ads appearing in ChatGPT, purpose-built extensions popped up to specifically block ads there. Existing ad blocking tools made sure to block those ads as well. It took hours and days, not weeks or months, for users to act.

Lesson two: the arms race is unwinnable, both legal and technical

The predictable response is, of course, to sue ad blockers. For some publishers, the idea that users take agency over their own screens was not acceptable, so they took ad blocker developers to court. Almost all of the legal attempts failed. With one case still pending at the German courts, China is the only country in the world that bans ad blockers.

Another tactic is technology to fight ad blockers. Platforms tried different circumvention methods to deliver ads to users with ad blockers installed. This prompted their developers to update their techniques to prevent websites from bypassing. Every technical escalation met a counter-escalation and the sum of a lot of engineering efforts on both sides is that, today, over a billion people still block ads.

AI platforms hold one card their predecessors didn't, though: integrating advertising directly into the generated answers, instead of showing labelled ads. Researchers from the University of Michigan tried exactly this and the results should worry everyone involved: users largely don't expect advertising inside a generated answer and frequently fail to notice it.

At the same time, 63% of consumers say ads decrease their trust in AI outputs, according to Ipsos. The format that works best technically is the one users are least equipped to detect, and least willing to trust. Which brings us to the most important lesson.

Lesson three: trust is the actual product

Search results pages survived the ad blocking wars because users had ten blue links to compare with the sponsored content and two decades of learning what’s paid and what’s organic. AI assistants, on the other hand, offer one authoritative answer, with no other content or adjacent organic results to check against.

As soon as users suspect that answers might be influenced by a sponsor, these doubts are attached to all answers and trust erodes.

Cory Doctorow coined the term “enshittification”, which the American Dialect Society crowned as its Word of the Year in 2023. Platforms start out by doing good to their users and then degrade their experience to make advertisers happy.

And compared to tracking-based advertising in the open web, AI assistants know so much more about you that they can sell to advertisers: your health worries, your finances, your political view, your job struggles and everything else these tools picked up from you, stitched into one persistent memory of who you are.

It’s the most revealing personal record people have ever handed to software, combined with the drive from digital advertising to always collect more data and more data for targeting. The table is set for the next wave of enshittification.

Lesson four: the ad blocking wars ended in a treaty, not a victory

There was no final winner and not one technical solution that settled ad blocking in the open web. There was a rough settlement: standards for ad formats many users tolerate, transparency about what is paid for, meaningful user control.

Not everyone loves that settlement and we still see platforms trying to circumvent ad blockers and vice versa. But the open web still has a functioning ad economy and users still have, mostly, a functioning veto.

Getting there took more than a decade and the lessons are written down. Which road to take is now a genuine choice and it sits with the AI companies themselves.

We've featured the best CDN providers.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit

Maybe money does not explain everything, but it explains a lot. Goldman Sachs estimates the AI build-out will require roughly $7.6 trillion in capital over the next five years across compute, data centers and power. It’s unlikely that subscriptions alone will cover the bill, which is why ads are coming to AI.

OpenAI started rolling out ads in some ChatGPT tiers this year and forecasts $2.5 billion in ad revenue in 2026 and $100 billion by 2030. At an ad tech event this summer, OpenAI's chief revenue officer Denise Dresser dropped the hedging and said the company is "clearly in the advertising business now."

Perplexity, on the other hand, shut down its ad business entirely over user trust. Similar products, same opportunity, opposite conclusions.

This conflict is not new and the open web fought this battle 15 years ago, when ad blocking became mainstream.

Now, the AI industry is on track to repeat every mistake of the so-called ad blocking wars, unless it studies them instead.

Lesson one: users revolt faster than big platforms think

When ad formats became more intrusive - because again, money! - the industry assumed users would just accept a bad experience. Instead, users fought back and installed software to get rid of aggressive online ads. In what technology commentator Doc Searls called “the biggest boycott in world history”, ad blocking went from being a niche tool to a mainstream product.

And we see the same dynamic playing out again. Within weeks of sponsored ads appearing in ChatGPT, purpose-built extensions popped up to specifically block ads there. Existing ad blocking tools made sure to block those ads as well. It took hours and days, not weeks or months, for users to act.

Lesson two: the arms race is unwinnable, both legal and technical

The predictable response is, of course, to sue ad blockers. For some publishers, the idea that users take agency over their own screens was not acceptable, so they took ad blocker developers to court. Almost all of the legal attempts failed. With one case still pending at the German courts, China is the only country in the world that bans ad blockers.

Another tactic is technology to fight ad blockers. Platforms tried different circumvention methods to deliver ads to users with ad blockers installed. This prompted their developers to update their techniques to prevent websites from bypassing. Every technical escalation met a counter-escalation and the sum of a lot of engineering efforts on both sides is that, today, over a billion people still block ads.

AI platforms hold one card their predecessors didn't, though: integrating advertising directly into the generated answers, instead of showing labelled ads. Researchers from the University of Michigan tried exactly this and the results should worry everyone involved: users largely don't expect advertising inside a generated answer and frequently fail to notice it.

At the same time, 63% of consumers say ads decrease their trust in AI outputs, according to Ipsos. The format that works best technically is the one users are least equipped to detect, and least willing to trust. Which brings us to the most important lesson.

Lesson three: trust is the actual product

Search results pages survived the ad blocking wars because users had ten blue links to compare with the sponsored content and two decades of learning what’s paid and what’s organic. AI assistants, on the other hand, offer one authoritative answer, with no other content or adjacent organic results to check against.

As soon as users suspect that answers might be influenced by a sponsor, these doubts are attached to all answers and trust erodes.

Cory Doctorow coined the term “enshittification”, which the American Dialect Society crowned as its Word of the Year in 2023. Platforms start out by doing good to their users and then degrade their experience to make advertisers happy.

And compared to tracking-based advertising in the open web, AI assistants know so much more about you that they can sell to advertisers: your health worries, your finances, your political view, your job struggles and everything else these tools picked up from you, stitched into one persistent memory of who you are.

It’s the most revealing personal record people have ever handed to software, combined with the drive from digital advertising to always collect more data and more data for targeting. The table is set for the next wave of enshittification.

Lesson four: the ad blocking wars ended in a treaty, not a victory

There was no final winner and not one technical solution that settled ad blocking in the open web. There was a rough settlement: standards for ad formats many users tolerate, transparency about what is paid for, meaningful user control.

Not everyone loves that settlement and we still see platforms trying to circumvent ad blockers and vice versa. But the open web still has a functioning ad economy and users still have, mostly, a functioning veto.

Getting there took more than a decade and the lessons are written down. Which road to take is now a genuine choice and it sits with the AI companies themselves.

We've featured the best CDN providers.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit

When you spend much of the day at a desk, investing in an ergonomic office chair can be money well spent. And one of our favorites is the Doro C300 Pro V2, currently $400 (was $500) at Sihoo in the Labor Day sale when you combine codes BTS20 and WELCOME10.

Testing this desk chair, we found it comfortable, highly adjustable, and very well built. Yes, it's still a considerable investment, but you're getting a high-end ergonomic design with adjustable seat depth, 8D armrests, adaptive lumbar support and a 135-degree recline.

The breathable mesh construction will also help keep you comfortable, and the wide adjustable headrest and weight-responsive recline system provide additional ways to tailor the chair to how you prefer to sit. In the UK, the Doro C300 V2 is down to £410 (was £460) at Sihoo with code Sale20 as part of the back-to-school sale.

Doro C300 Pro V2 Ergonomic Office Chair: was $499.99 now $399.99
The chair combines breathable mesh construction with self-adaptive lumbar support, a flexible backrest, adjustable seat depth, 8D armrests and a wide 3D headrest. You can recline to 105, 125 or 135 degrees, with the resistance automatically responding to your weight. Use coupon codes BTS20 and WELCOME10 to get the cheapest price.

In the UK: now £410 (was £460) at Sihoo with code Sale20View Deal

Why we recommend it

One of the more interesting things about Sihoo's Doro C300 Pro V2 is that many of its ergonomic features aren't simply fixed supports that you adjust once and forget about. The backrest and lumbar support are designed to respond as you lean forward, recline or otherwise change position.

In his 4-star review, our chair expert Collin called it a "robust ergonomic office chair with more features than you probably know what to do with."

He added: "This is a spectacular chair when you sit in it. It doesn't blend in and takes up a lot of space. So, if you’re the kind of person who likes that, this may be a dream come true."

Should you buy it?

✅ Buy the Sihoo Doro C300 Pro V2 if...

If you spend much of your working day sitting at a computer, the Doro C300 Pro V2 has an impressive range of ergonomic features. Adaptive lumbar and back support, adjustable seat depth and a weight-responsive recline system allow the chair to respond to your position rather than relying entirely on fixed settings.

The 8D armrests, wide 3D headrest and three recline positions give you plenty of scope to find the perfect setup.

❌ Skip the Sihoo Doro C300 Pro V2 if...

If you only spend a few hours a week at a desk, it's hard to justify paying $400 for features you may rarely appreciate. A simpler ergonomic chair could provide everything you need while costing considerably less.

The Doro C300 Pro V2 is also relatively large and relies heavily on supportive mesh rather than the thick padding found on more traditional executive-style chairs.

Price context & historical value

The Sihoo Doro C300 Pro V2 is currently $409.99 at Sihoo, down from the usual $499.99. That's a pretty good deal, but if you use coupon codes BTS20 and WELCOME10 you can slash the price down to a penny under $400.

The chair is also on sale at Amazon at the moment, but it's $449.98 there.

The Catch: What to know before you buy

The obvious drawback is the price. $399.99 is still a lot to spend on an office chair. There's also an important point about highly adjustable ergonomic chairs: more adjustment doesn't automatically mean greater comfort for everyone. The Doro C300 Pro V2 offers plenty of ways to tailor your seating position, but personal preference, height and body shape can make a considerable difference to how comfortable any chair feels.

Its 28.74-inch width also means this chair might be a bit large for anyone working in a more compact office space.

• US: Save up to 50% in Sihoo's Labor Day sale
• UK: Get up to 50% off in the back-to-school sale at Sihoo

When you spend much of the day at a desk, investing in an ergonomic office chair can be money well spent. And one of our favorites is the Doro C300 Pro V2, currently $400 (was $500) at Sihoo in the Labor Day sale when you combine codes BTS20 and WELCOME10.

Testing this desk chair, we found it comfortable, highly adjustable, and very well built. Yes, it's still a considerable investment, but you're getting a high-end ergonomic design with adjustable seat depth, 8D armrests, adaptive lumbar support and a 135-degree recline.

The breathable mesh construction will also help keep you comfortable, and the wide adjustable headrest and weight-responsive recline system provide additional ways to tailor the chair to how you prefer to sit. In the UK, the Doro C300 V2 is down to £410 (was £460) at Sihoo with code Sale20 as part of the back-to-school sale.

Doro C300 Pro V2 Ergonomic Office Chair: was $499.99 now $399.99
The chair combines breathable mesh construction with self-adaptive lumbar support, a flexible backrest, adjustable seat depth, 8D armrests and a wide 3D headrest. You can recline to 105, 125 or 135 degrees, with the resistance automatically responding to your weight. Use coupon codes BTS20 and WELCOME10 to get the cheapest price.

In the UK: now £410 (was £460) at Sihoo with code Sale20View Deal

Why we recommend it

One of the more interesting things about Sihoo's Doro C300 Pro V2 is that many of its ergonomic features aren't simply fixed supports that you adjust once and forget about. The backrest and lumbar support are designed to respond as you lean forward, recline or otherwise change position.

In his 4-star review, our chair expert Collin called it a "robust ergonomic office chair with more features than you probably know what to do with."

He added: "This is a spectacular chair when you sit in it. It doesn't blend in and takes up a lot of space. So, if you’re the kind of person who likes that, this may be a dream come true."

Should you buy it?

✅ Buy the Sihoo Doro C300 Pro V2 if...

If you spend much of your working day sitting at a computer, the Doro C300 Pro V2 has an impressive range of ergonomic features. Adaptive lumbar and back support, adjustable seat depth and a weight-responsive recline system allow the chair to respond to your position rather than relying entirely on fixed settings.

The 8D armrests, wide 3D headrest and three recline positions give you plenty of scope to find the perfect setup.

❌ Skip the Sihoo Doro C300 Pro V2 if...

If you only spend a few hours a week at a desk, it's hard to justify paying $400 for features you may rarely appreciate. A simpler ergonomic chair could provide everything you need while costing considerably less.

The Doro C300 Pro V2 is also relatively large and relies heavily on supportive mesh rather than the thick padding found on more traditional executive-style chairs.

Price context & historical value

The Sihoo Doro C300 Pro V2 is currently $409.99 at Sihoo, down from the usual $499.99. That's a pretty good deal, but if you use coupon codes BTS20 and WELCOME10 you can slash the price down to a penny under $400.

The chair is also on sale at Amazon at the moment, but it's $449.98 there.

The Catch: What to know before you buy

The obvious drawback is the price. $399.99 is still a lot to spend on an office chair. There's also an important point about highly adjustable ergonomic chairs: more adjustment doesn't automatically mean greater comfort for everyone. The Doro C300 Pro V2 offers plenty of ways to tailor your seating position, but personal preference, height and body shape can make a considerable difference to how comfortable any chair feels.

Its 28.74-inch width also means this chair might be a bit large for anyone working in a more compact office space.

• US: Save up to 50% in Sihoo's Labor Day sale
• UK: Get up to 50% off in the back-to-school sale at Sihoo

When you run AI agents like ChatGPT Work or Microsoft Copilot, your files and prompts all pass through third-party servers that you’re not in control of. Essentially, you’re paying a subscription fee for the privilege of letting someone else handle your data, when you could be running the same tasks locally from your own hardware.

For professionals and teams that work in regulated industries or need to ensure strict local compliance, this can be an ongoing challenge. Even when you’re opting out of model training in ChatGPT or Claude, you’re trusting them to honor that agreement with very little legal recourse in case they don’t.

However, self-hosting your own agents is not as complicated as it used to be. You can easily set up a private OpenClaw environment on your local workstation, VPS, or NAS. Then if you want, you can go the extra mile by layering your OpenClaw setup with Open WebUI’s polished visual interface. I’ve run this setup myself on my home computer, and I’ll show you how easily you can do it too.

What you'll need before we start

• A local desktop, virtual machine, VPS, or NAS device running Linux. It should have at least 4GB of RAM and 20GB of free storage space.
• Sudo access to your Linux operating system.
• Node.js 22 or later installed on your OS (Node.js 24 is the recommended version).
• An API key for your preferred LLM vendor, such as OpenAI or Anthropic.
• Or if you have the GPU for it, you can run an open-weight LLM like Llama 3.3 locally on your system with Ollama (optional).

(Image credit: Openclaw/Edited with Gemini )Step-by-step installation guide for self-hosting your own AI agent Step 1: Install OpenClaw

OpenClaw is an open-source autonomous agent framework that can run shell commands, browse the web, read/write files, and connect to messaging apps using your local infrastructure. You can install it using the official script.

curl -fsSL https://openclaw.ai/install.sh | bash

Once installed, run the onboarding wizard. This is where you'll connect your model provider's API key and give your OpenClaw agent a name.

openclaw onboard

Follow the on-screen instructions to configure a default AI model and paste in your API key after picking. You can also choose to run an open-weight AI model like Llama 3.3 or Qwen 3.7 locally on your hardware, by downloading the model repo from an online database like Hugging Face. The wizard writes its output to a configuration file at ~/.openclaw/openclaw.json, which you'll need to edit again in the next step.

Step 2: Enable the API endpoint

Open WebUI talks to OpenClaw through an OpenAI-compatible endpoint, but that endpoint is switched off by default for security reasons. Open the configuration file in a text editor using the nano command.

nano ~/.openclaw/openclaw.json

Now add or update the following block so the API endpoint is enabled.

json{ "gateway": { "http": { "endpoints": { "chatCompletions": { "enabled": true } } } } }

Save the file and close the editor. If the gateway is already running, you'll need to restart it for the change to take effect.

Step 3: Start the gateway

With the endpoint enabled, we need to start the gateway process.

openclaw gateway

The gateway starts on port 18789 by default. Confirm it's running with the built-in status check.

openclaw gateway status

(Image credit: Openclaw/Edited with Gemini )Step 4: Install Docker and run Open WebUI

Open WebUI offers a polished visual interface that makes it easier to create and run OpenClaw agents without interacting with the CLI in terminal every time. Open WebUI recommends installing it as a containerized Docker application, so we first need to install Docker Engine and Docker Compose.

Before we do anything else, open a terminal window and update your package index.

sudo apt-get updatesudo apt-get install -y ca-certificates curl

Add Docker's official GPG key, which lets your system verify that the packages actually came from Docker.

sudo install -m 0755 -d /etc/apt/keyringssudo curl -fsSL https://download.docker.com/linux/ubuntu/gpg -o /etc/apt/keyrings/docker.ascsudo chmod a+r /etc/apt/keyrings/docker.asc

Now register Docker's repository with apt.

echo \ "deb [arch=$(dpkg --print-architecture) signed-by=/etc/apt/keyrings/docker.asc] https://download.docker.com/linux/ubuntu \ $(. /etc/os-release && echo "$VERSION_CODENAME") stable" | sudo tee /etc/apt/sources.list.d/docker.list > /dev/nullsudo apt-get update

Install Docker Engine and Docker Compose.

sudo apt-get install -y docker-ce docker-ce-cli containerd.io docker-buildx-plugin docker-compose-plugin

Now run Open WebUI as a container.

docker run -d \ --name open-webui \ --restart unless-stopped \ -p 3000:8080 \ --add-host=host.docker.internal:host-gateway \ -v open-webui-data:/app/backend/data \ ghcr.io/open-webui/open-webui:main

Note: Because Open WebUI runs in a container and OpenClaw runs directly on the host, Open WebUI can't reach the gateway using localhost. Using the --add-host flag here lets it reach the host machine instead, where the OpenClaw gateway is listening on port 18789.

Step 5: Connect Open WebUI to your OpenClaw agent

Open a browser and navigate to http://your-server-ip:3000. Create an admin account on first launch, since the very first user to sign up becomes the administrator.

Once you're logged in, head to the connections settings.

• Go to Admin Settings > Connections > OpenAI.
• Click Add Connection.
• For the URL, enter http://host.docker.internal:18789/v1.
• For the API key, paste your OpenClaw gateway bearer token, which you'll find in ~/.openclaw/openclaw.json under the gateway authentication section.
• Click the checkmark to verify the connection, then save.

Your OpenClaw agent should now appear as a selectable model in Open WebUI's chat window. Pick it from the dropdown and send a test message to confirm everything is wired up correctly.

(Image credit: Openclaw/Edited with Gemini )What to do now that you’re all set up?

With your OpenClaw setup complete and the UI functional, you can now start exploring the kind of automated workflows you can run with this system. If you’re flummoxed by all the options, here are some pointers on what you could look into first:

• OpenClaw has a community-driven Skills library that features tons of working automations for everything from GitHub repo maintenance to data visualization to vibe coding. Take a look at the ClawHub to get started.
• Apart from Open WebUI, OpenClaw can also be integrated with your favorite communication apps like Slack, Telegram, Discord, or WhatsApp. Or better yet, you can use the built-in integration tools offered by OpenWebUI to set up a chatbot for Telegram, WhatsApp, Signal, and others. Just navigate to Settings > Admin > Bots in Open WebUI to configure chatbots using API tokens.
• If you want to run automated workflows on repeat, OpenClaw lets you set up cron jobs that trigger automatically according to a set frequency. View the OpenClaw documentation for details on how to set these up.
• If you’re running this setup on a personal workstation, make sure to back up all your important files and log out of active online accounts that you don’t want your agents to have access to. OpenClaw has system-wide privileges enabled by default, so it can read and write any file or execute tasks on its own, which can be a problem if you have sensitive data on the device. This is also why setting up OpenClaw on a VPS server is the recommended method for data isolation.

FAQsDo I need coding experience for OpenClaw?

No, you don’t need to know any programming languages for OpenClaw, although some basic familiarity with command-line interfaces does help during setup. Once the first-time setup is complete, though, you can simply rely on Open WebUI’s chat-based interface to configure and run OpenClaw agents.

Is this really more private? Does my data never leave the system?

It depends on which options you chose when following this tutorial. If you’re connecting OpenClaw a cloud-hosted AI model’s private API, the model provider will have access to any data you process through the LLM. However, running an open-weight model locally on your system hardware keeps everything confined to your own environment with no cloud involvement.

Are there any subscription fees or costs associated with this setup?

OpenClaw and Open WebUI are both free to use. If you use a cloud-based AI model to run OpenClaw, however, you’ll still need to pay for API credits based on your LLM’s token consumption. This is still much more cost-effective than a fully managed AI agent platform, which will bill you separately for each automated job you run.

When you run AI agents like ChatGPT Work or Microsoft Copilot, your files and prompts all pass through third-party servers that you’re not in control of. Essentially, you’re paying a subscription fee for the privilege of letting someone else handle your data, when you could be running the same tasks locally from your own hardware.

For professionals and teams that work in regulated industries or need to ensure strict local compliance, this can be an ongoing challenge. Even when you’re opting out of model training in ChatGPT or Claude, you’re trusting them to honor that agreement with very little legal recourse in case they don’t.

However, self-hosting your own agents is not as complicated as it used to be. You can easily set up a private OpenClaw environment on your local workstation, VPS, or NAS. Then if you want, you can go the extra mile by layering your OpenClaw setup with Open WebUI’s polished visual interface. I’ve run this setup myself on my home computer, and I’ll show you how easily you can do it too.

What you'll need before we start

• A local desktop, virtual machine, VPS, or NAS device running Linux. It should have at least 4GB of RAM and 20GB of free storage space.
• Sudo access to your Linux operating system.
• Node.js 22 or later installed on your OS (Node.js 24 is the recommended version).
• An API key for your preferred LLM vendor, such as OpenAI or Anthropic.
• Or if you have the GPU for it, you can run an open-weight LLM like Llama 3.3 locally on your system with Ollama (optional).

(Image credit: Openclaw/Edited with Gemini )Step-by-step installation guide for self-hosting your own AI agent Step 1: Install OpenClaw

OpenClaw is an open-source autonomous agent framework that can run shell commands, browse the web, read/write files, and connect to messaging apps using your local infrastructure. You can install it using the official script.

curl -fsSL https://openclaw.ai/install.sh | bash

Once installed, run the onboarding wizard. This is where you'll connect your model provider's API key and give your OpenClaw agent a name.

openclaw onboard

Follow the on-screen instructions to configure a default AI model and paste in your API key after picking. You can also choose to run an open-weight AI model like Llama 3.3 or Qwen 3.7 locally on your hardware, by downloading the model repo from an online database like Hugging Face. The wizard writes its output to a configuration file at ~/.openclaw/openclaw.json, which you'll need to edit again in the next step.

Step 2: Enable the API endpoint

Open WebUI talks to OpenClaw through an OpenAI-compatible endpoint, but that endpoint is switched off by default for security reasons. Open the configuration file in a text editor using the nano command.

nano ~/.openclaw/openclaw.json

Now add or update the following block so the API endpoint is enabled.

json{ "gateway": { "http": { "endpoints": { "chatCompletions": { "enabled": true } } } } }

Save the file and close the editor. If the gateway is already running, you'll need to restart it for the change to take effect.

Step 3: Start the gateway

With the endpoint enabled, we need to start the gateway process.

openclaw gateway

The gateway starts on port 18789 by default. Confirm it's running with the built-in status check.

openclaw gateway status

(Image credit: Openclaw/Edited with Gemini )Step 4: Install Docker and run Open WebUI

Open WebUI offers a polished visual interface that makes it easier to create and run OpenClaw agents without interacting with the CLI in terminal every time. Open WebUI recommends installing it as a containerized Docker application, so we first need to install Docker Engine and Docker Compose.

Before we do anything else, open a terminal window and update your package index.

sudo apt-get updatesudo apt-get install -y ca-certificates curl

Add Docker's official GPG key, which lets your system verify that the packages actually came from Docker.

sudo install -m 0755 -d /etc/apt/keyringssudo curl -fsSL https://download.docker.com/linux/ubuntu/gpg -o /etc/apt/keyrings/docker.ascsudo chmod a+r /etc/apt/keyrings/docker.asc

Now register Docker's repository with apt.

echo \ "deb [arch=$(dpkg --print-architecture) signed-by=/etc/apt/keyrings/docker.asc] https://download.docker.com/linux/ubuntu \ $(. /etc/os-release && echo "$VERSION_CODENAME") stable" | sudo tee /etc/apt/sources.list.d/docker.list > /dev/nullsudo apt-get update

Install Docker Engine and Docker Compose.

sudo apt-get install -y docker-ce docker-ce-cli containerd.io docker-buildx-plugin docker-compose-plugin

Now run Open WebUI as a container.

docker run -d \ --name open-webui \ --restart unless-stopped \ -p 3000:8080 \ --add-host=host.docker.internal:host-gateway \ -v open-webui-data:/app/backend/data \ ghcr.io/open-webui/open-webui:main

Note: Because Open WebUI runs in a container and OpenClaw runs directly on the host, Open WebUI can't reach the gateway using localhost. Using the --add-host flag here lets it reach the host machine instead, where the OpenClaw gateway is listening on port 18789.

Step 5: Connect Open WebUI to your OpenClaw agent

Open a browser and navigate to http://your-server-ip:3000. Create an admin account on first launch, since the very first user to sign up becomes the administrator.

Once you're logged in, head to the connections settings.

• Go to Admin Settings > Connections > OpenAI.
• Click Add Connection.
• For the URL, enter http://host.docker.internal:18789/v1.
• For the API key, paste your OpenClaw gateway bearer token, which you'll find in ~/.openclaw/openclaw.json under the gateway authentication section.
• Click the checkmark to verify the connection, then save.

Your OpenClaw agent should now appear as a selectable model in Open WebUI's chat window. Pick it from the dropdown and send a test message to confirm everything is wired up correctly.

(Image credit: Openclaw/Edited with Gemini )What to do now that you’re all set up?

With your OpenClaw setup complete and the UI functional, you can now start exploring the kind of automated workflows you can run with this system. If you’re flummoxed by all the options, here are some pointers on what you could look into first:

• OpenClaw has a community-driven Skills library that features tons of working automations for everything from GitHub repo maintenance to data visualization to vibe coding. Take a look at the ClawHub to get started.
• Apart from Open WebUI, OpenClaw can also be integrated with your favorite communication apps like Slack, Telegram, Discord, or WhatsApp. Or better yet, you can use the built-in integration tools offered by OpenWebUI to set up a chatbot for Telegram, WhatsApp, Signal, and others. Just navigate to Settings > Admin > Bots in Open WebUI to configure chatbots using API tokens.
• If you want to run automated workflows on repeat, OpenClaw lets you set up cron jobs that trigger automatically according to a set frequency. View the OpenClaw documentation for details on how to set these up.
• If you’re running this setup on a personal workstation, make sure to back up all your important files and log out of active online accounts that you don’t want your agents to have access to. OpenClaw has system-wide privileges enabled by default, so it can read and write any file or execute tasks on its own, which can be a problem if you have sensitive data on the device. This is also why setting up OpenClaw on a VPS server is the recommended method for data isolation.

FAQsDo I need coding experience for OpenClaw?

No, you don’t need to know any programming languages for OpenClaw, although some basic familiarity with command-line interfaces does help during setup. Once the first-time setup is complete, though, you can simply rely on Open WebUI’s chat-based interface to configure and run OpenClaw agents.

Is this really more private? Does my data never leave the system?

It depends on which options you chose when following this tutorial. If you’re connecting OpenClaw a cloud-hosted AI model’s private API, the model provider will have access to any data you process through the LLM. However, running an open-weight model locally on your system hardware keeps everything confined to your own environment with no cloud involvement.

Are there any subscription fees or costs associated with this setup?

OpenClaw and Open WebUI are both free to use. If you use a cloud-based AI model to run OpenClaw, however, you’ll still need to pay for API credits based on your LLM’s token consumption. This is still much more cost-effective than a fully managed AI agent platform, which will bill you separately for each automated job you run.

• Dutch investigation and French complaint accuse Uber of banning drivers via algorithms
• 171 drivers affected on the basis of driver behavior and customer feedback
• Uber says humans were involved in permanent account deactivations

The Dutch Data Protection Authority (AP) has issued Uber with a hefty €825 million ($962 million) fine over the way it used automated systems to suspend driver accounts within Europe.

Because losing access to Uber can immediately prevent drivers from earning money through the platform, regulators deemed that the case should fall under Article 22 of the EU GDPR, which restricts automated-only decision-making when it comes to having significant effects on individuals.

Ultimately, it means this is now the second-largest GDPR fine ever to have been issued, falling short of Meta's 2023 €1.2 billion fine.

Uber hit with €825 million GDPR fine

Though the company was finally hit with the fine in 2026, it relates to business practices between 2018 and 2022 when Uber used software to assess driver behavior and customer feedback before automatically restricting suspicious accounts.

Uber's systems reportedly flagged drivers for taking unnecessary detours to inflate customer prices, and accepting journeys without intending to complete them – perfectly reasonable reasons to be flagged.

But Dutch regulators were more concerned with the fact that decisions were being made entirely by computers and algorithms without human approval, leading drivers to potentially lose their livelihoods. "A computer should not make decisions on its own that have (such) major consequences," regulators said.

Netherlands led the investigation because Uber's main European establishment is based in the country, but the case stems back to an earlier 2020 complaint by French human rights organization La Ligue des Droits de l'Homme (LDH), representing 171 Uber drivers.

Uber argues that fraud-related suspensions were generally short-lived, and that permanent account deactivations were not made without human review, thus it intends to appeal. "We strongly disagree with this decision and disproportionate fine," the company said in a statement (via Reuters).

• Dutch investigation and French complaint accuse Uber of banning drivers via algorithms
• 171 drivers affected on the basis of driver behavior and customer feedback
• Uber says humans were involved in permanent account deactivations

The Dutch Data Protection Authority (AP) has issued Uber with a hefty €825 million ($962 million) fine over the way it used automated systems to suspend driver accounts within Europe.

Because losing access to Uber can immediately prevent drivers from earning money through the platform, regulators deemed that the case should fall under Article 22 of the EU GDPR, which restricts automated-only decision-making when it comes to having significant effects on individuals.

Ultimately, it means this is now the second-largest GDPR fine ever to have been issued, falling short of Meta's 2023 €1.2 billion fine.

Uber hit with €825 million GDPR fine

Though the company was finally hit with the fine in 2026, it relates to business practices between 2018 and 2022 when Uber used software to assess driver behavior and customer feedback before automatically restricting suspicious accounts.

Uber's systems reportedly flagged drivers for taking unnecessary detours to inflate customer prices, and accepting journeys without intending to complete them – perfectly reasonable reasons to be flagged.

But Dutch regulators were more concerned with the fact that decisions were being made entirely by computers and algorithms without human approval, leading drivers to potentially lose their livelihoods. "A computer should not make decisions on its own that have (such) major consequences," regulators said.

Netherlands led the investigation because Uber's main European establishment is based in the country, but the case stems back to an earlier 2020 complaint by French human rights organization La Ligue des Droits de l'Homme (LDH), representing 171 Uber drivers.

Uber argues that fraud-related suspensions were generally short-lived, and that permanent account deactivations were not made without human review, thus it intends to appeal. "We strongly disagree with this decision and disproportionate fine," the company said in a statement (via Reuters).

There is a point in many endpoint projects when the new environment appears to be ready.

Devices are enrolled, policies are in place and physical endpoints and virtual desktops—including Cloud PCs—can be provisioned.

From an infrastructure perspective, most of the difficult work seems to have been done.

The applications often tell a different story.

Moving an older application onto a new platform does not alter its underlying behavior.

It may still rely on ageing components and expect administrator access to write files and registry entries across the device. Its update process may also be largely manual or poorly documented.

A well-managed endpoint is valuable, but it cannot make a difficult application easier to maintain simply by hosting it.

A successful deployment can hide the real work

Modern management platforms have made it much quicker to enroll devices, assign software and apply policies across an organization.

Cloud services and virtual desktops have also taken much of the effort out of provisioning the underlying environment. These are genuine advances, but they can create a false sense of completion. An application may open in the new environment and appear ready for production.

The problems tend to emerge later. An update arrives and the packaging process has to be recreated. An operating system change exposes a dependency that was never recorded. A certificate expires, or a new release behaves differently from the version originally tested.

None of these issues are solved by the fact that the application was delivered through a modern platform. Deployment gets an application to the user; modernization determines whether IT can continue to manage it once the project team has moved on.

Old assumptions follow applications into new environments

Enterprise application estates tend to grow gradually. Software is added over many years, often under different teams and different versions of Windows. Some applications expect access to folders or registry locations that would not be considered appropriate today. Others rely on old runtimes, fixed paths or components installed so long ago that nobody is entirely sure why they are there. Accompanying IT documentation is rarely complete.

The original installer may have disappeared and the person who understood the application may no longer work for the organization, yet the software itself may remain essential to a finance team, production line or clinical process.

Moving to modern endpoint management such as Intune, a Cloud PC or a new virtual desktop environment does not remove those assumptions. In some cases, it exposes them because modern environments place tighter limits on what software is allowed to change. A launch test will reveal only part of the picture.

IT management also needs to know how the application installs, what it depends on, which parts of the system it expects to access and how it is updated and removed. Age alone is not a reason to replace an application. Many older line-of-business tools remain useful and business-critical. The aim should be to understand the software well enough to manage it properly, rather than carrying its old behavior into a new environment without examining it.

Modernization does not have to mean rewriting software

Rewriting every legacy application would be impractical for most businesses and may deliver little value where the existing software still performs its function well. A better test is whether the application can be supported throughout the rest of its working life. That requires a clear view of its files, services, dependencies and interaction with Windows. Once those details are known, compatibility problems can be addressed deliberately rather than discovered by users after deployment.

Packaging is part of that work, although it is not the whole answer. On Windows, formats such as MSIX can give an application a clearer identity, require digital signing and separate more of its activity from the underlying operating system. They can also support cleaner removal and more controlled updates. MSIX can provide a more secure and manageable foundation, although applications with legacy middleware, undocumented dependencies or highly specialized configurations may still require additional work.

This is why organizations need to identify complex applications early. Testing must also reflect real use: opening the main window does not show whether employees can complete the tasks they rely on, whether the next update will install successfully or whether the application will continue working when another part of the environment changes.

The scope becomes clearer when organizations view the modern endpoint as a set of interconnected components. A desktop comprises three core elements: the operating system, the user profile, and applications. Hardware is the fourth component, but the same principles apply whether the machine is physical or virtual.

The applications work is therefore a project in its own right. Applications must be brought under control, secured, and kept up to date and delivered to the right users, with the control and compliance an organization requires.

Bring applications into the project sooner

Applications are often treated as the last stage of an endpoint program. The platform is selected, the environment is built and only then does the organization begin working through the software estate. By that point, timelines have been agreed and assumptions about compatibility may already be built into the plan. Difficult applications emerge as late exceptions, when there is the least room to deal with them.

Bringing application assessment forward gives IT a clearer view of what it is migrating. Teams can establish which applications are in use, which are business-critical and which depend on technology that may not be available in the new environment. They can also separate routine applications from those requiring deeper technical or business knowledge.

I have worked in this industry since the late 1990s and have seen enterprise desktops move from early thin-client systems to virtual desktops and, more recently, cloud-hosted PCs. The infrastructure has improved enormously, but the questions around applications have barely changed.

A modern endpoint is an important step, but the real measure of modernization is whether the organization can understand, update and support the applications running on it.

We've rounded up the best endpoint protection software suites.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit

There is a point in many endpoint projects when the new environment appears to be ready.

Devices are enrolled, policies are in place and physical endpoints and virtual desktops—including Cloud PCs—can be provisioned.

From an infrastructure perspective, most of the difficult work seems to have been done.

The applications often tell a different story.

Moving an older application onto a new platform does not alter its underlying behavior.

It may still rely on ageing components and expect administrator access to write files and registry entries across the device. Its update process may also be largely manual or poorly documented.

A well-managed endpoint is valuable, but it cannot make a difficult application easier to maintain simply by hosting it.

A successful deployment can hide the real work

Modern management platforms have made it much quicker to enroll devices, assign software and apply policies across an organization.

Cloud services and virtual desktops have also taken much of the effort out of provisioning the underlying environment. These are genuine advances, but they can create a false sense of completion. An application may open in the new environment and appear ready for production.

The problems tend to emerge later. An update arrives and the packaging process has to be recreated. An operating system change exposes a dependency that was never recorded. A certificate expires, or a new release behaves differently from the version originally tested.

None of these issues are solved by the fact that the application was delivered through a modern platform. Deployment gets an application to the user; modernization determines whether IT can continue to manage it once the project team has moved on.

Old assumptions follow applications into new environments

Enterprise application estates tend to grow gradually. Software is added over many years, often under different teams and different versions of Windows. Some applications expect access to folders or registry locations that would not be considered appropriate today. Others rely on old runtimes, fixed paths or components installed so long ago that nobody is entirely sure why they are there. Accompanying IT documentation is rarely complete.

The original installer may have disappeared and the person who understood the application may no longer work for the organization, yet the software itself may remain essential to a finance team, production line or clinical process.

Moving to modern endpoint management such as Intune, a Cloud PC or a new virtual desktop environment does not remove those assumptions. In some cases, it exposes them because modern environments place tighter limits on what software is allowed to change. A launch test will reveal only part of the picture.

IT management also needs to know how the application installs, what it depends on, which parts of the system it expects to access and how it is updated and removed. Age alone is not a reason to replace an application. Many older line-of-business tools remain useful and business-critical. The aim should be to understand the software well enough to manage it properly, rather than carrying its old behavior into a new environment without examining it.

Modernization does not have to mean rewriting software

Rewriting every legacy application would be impractical for most businesses and may deliver little value where the existing software still performs its function well. A better test is whether the application can be supported throughout the rest of its working life. That requires a clear view of its files, services, dependencies and interaction with Windows. Once those details are known, compatibility problems can be addressed deliberately rather than discovered by users after deployment.

Packaging is part of that work, although it is not the whole answer. On Windows, formats such as MSIX can give an application a clearer identity, require digital signing and separate more of its activity from the underlying operating system. They can also support cleaner removal and more controlled updates. MSIX can provide a more secure and manageable foundation, although applications with legacy middleware, undocumented dependencies or highly specialized configurations may still require additional work.

This is why organizations need to identify complex applications early. Testing must also reflect real use: opening the main window does not show whether employees can complete the tasks they rely on, whether the next update will install successfully or whether the application will continue working when another part of the environment changes.

The scope becomes clearer when organizations view the modern endpoint as a set of interconnected components. A desktop comprises three core elements: the operating system, the user profile, and applications. Hardware is the fourth component, but the same principles apply whether the machine is physical or virtual.

The applications work is therefore a project in its own right. Applications must be brought under control, secured, and kept up to date and delivered to the right users, with the control and compliance an organization requires.

Bring applications into the project sooner

Applications are often treated as the last stage of an endpoint program. The platform is selected, the environment is built and only then does the organization begin working through the software estate. By that point, timelines have been agreed and assumptions about compatibility may already be built into the plan. Difficult applications emerge as late exceptions, when there is the least room to deal with them.

Bringing application assessment forward gives IT a clearer view of what it is migrating. Teams can establish which applications are in use, which are business-critical and which depend on technology that may not be available in the new environment. They can also separate routine applications from those requiring deeper technical or business knowledge.

I have worked in this industry since the late 1990s and have seen enterprise desktops move from early thin-client systems to virtual desktops and, more recently, cloud-hosted PCs. The infrastructure has improved enormously, but the questions around applications have barely changed.

A modern endpoint is an important step, but the real measure of modernization is whether the organization can understand, update and support the applications running on it.

We've rounded up the best endpoint protection software suites.

This article was produced as part of TechRadar Pro Perspectives, our channel to feature the best and brightest minds in the technology industry today.

The views expressed here are those of the author and are not necessarily those of TechRadarPro or Future plc. If you are interested in contributing find out more here: https://www.techradar.com/pro/perspectives-how-to-submit

• Philips Hue could be launching a light module that lets you connect your Nanoleaf Shapes to the Hue app
• You'll be able to integrate Nanoleaf lights into your Hue schedules and automations
• It's not Matter-over-Thread, but you might be able to use a Hue Bridge for third-party connectivity

We’re just one week from IFA 2026, and we already have our ears to the ground for all the pre-event announcements and rumors — and this latest Philips Hue leak is bound to make Nanoleaf users very happy.

As spotted by the outlet Smartlights, Philips is reportedly launching the Hue Wall Light Module next week, meaning that users will be able to integrate their Nanoleaf lights into the Philips Hue ecosystem — which would be a huge win, especially since Nanoleaf is an affordable alternative to the Philips Hue range of best smart lights.

The Hue Wall Light Module is said to support the Nanoleaf Shapes LED panels, including Mini Triangles, Triangles, and Hexagons, while older panels (formerly known as Aurora) won’t be supported. Philips’ new module is also expected to be announced on September 3, and will go on sale for 39.99 euros, with international prices expected to follow.

Essentially, the rumored Hue module is a small white adapter around 8cm that you can plug directly into the connectors in your Nanoleaf panels, allowing you to integrate up to 100 Nanoleaf Shapes panels into your Hue system. That said, the Hue Wall Light Module shouldn’t be confused with the existing Philips Hue Wall Switch Module, which is installed behind a standard wall switch.

Once your Nanoleaf panels are connected via Philips’ new module, you’ll be able to include them in automations, Hue scenes, and schedules in the Hue app — but whether Nanoleaf’s flagship lighting tools will be carried over is unclear.

(Image credit: Smartlight.de)

Similar to Philips Hue, Nanoleaf offers customizable color scenes and animations, but as far as the leaks go, the product images show the Nanoleaf integration working primarily with Hue scenes and effects. The leak shows at least five points in the color selection wheel, suggesting that there’ll be some kind of multicolor control function regardless, but we’ll have to see how this will work after the launch.

The other catch is that the Philips Hue Wall Light Module isn’t Matter-over-Thread, which, considering Nanoleaf’s devices are mainly Thread- and Matter-supported, is a bit of a departure. Instead, the module uses Zigbee and Bluetooth, the main connectivity protocols Philips has used for some time.

That said, using a Hue Bridge allows you to connect Philips Hue devices that are non-Matter to hubs such as the Apple HomePod, which is helpful if you use the best smart speakers to control your home lighting system — though we’ll have to wait and see how the new Philips Hue module handles third-party connectivity once it’s launched.

Despite these gray areas, the Nanoleaf-Hue integration is an unexpected yet possibly beneficial upgrade. When it comes to smart lighting brands, their respective devices tend to keep you locked into one ecosystem or the other, so having a simple but effective connectivity tool would give users much more flexibility.

• Philips Hue could be launching a light module that lets you connect your Nanoleaf Shapes to the Hue app
• You'll be able to integrate Nanoleaf lights into your Hue schedules and automations
• It's not Matter-over-Thread, but you might be able to use a Hue Bridge for third-party connectivity

We’re just one week from IFA 2026, and we already have our ears to the ground for all the pre-event announcements and rumors — and this latest Philips Hue leak is bound to make Nanoleaf users very happy.

As spotted by the outlet Smartlights, Philips is reportedly launching the Hue Wall Light Module next week, meaning that users will be able to integrate their Nanoleaf lights into the Philips Hue ecosystem — which would be a huge win, especially since Nanoleaf is an affordable alternative to the Philips Hue range of best smart lights.

The Hue Wall Light Module is said to support the Nanoleaf Shapes LED panels, including Mini Triangles, Triangles, and Hexagons, while older panels (formerly known as Aurora) won’t be supported. Philips’ new module is also expected to be announced on September 3, and will go on sale for 39.99 euros, with international prices expected to follow.

Essentially, the rumored Hue module is a small white adapter around 8cm that you can plug directly into the connectors in your Nanoleaf panels, allowing you to integrate up to 100 Nanoleaf Shapes panels into your Hue system. That said, the Hue Wall Light Module shouldn’t be confused with the existing Philips Hue Wall Switch Module, which is installed behind a standard wall switch.

Once your Nanoleaf panels are connected via Philips’ new module, you’ll be able to include them in automations, Hue scenes, and schedules in the Hue app — but whether Nanoleaf’s flagship lighting tools will be carried over is unclear.

(Image credit: Smartlight.de)

Similar to Philips Hue, Nanoleaf offers customizable color scenes and animations, but as far as the leaks go, the product images show the Nanoleaf integration working primarily with Hue scenes and effects. The leak shows at least five points in the color selection wheel, suggesting that there’ll be some kind of multicolor control function regardless, but we’ll have to see how this will work after the launch.

The other catch is that the Philips Hue Wall Light Module isn’t Matter-over-Thread, which, considering Nanoleaf’s devices are mainly Thread- and Matter-supported, is a bit of a departure. Instead, the module uses Zigbee and Bluetooth, the main connectivity protocols Philips has used for some time.

That said, using a Hue Bridge allows you to connect Philips Hue devices that are non-Matter to hubs such as the Apple HomePod, which is helpful if you use the best smart speakers to control your home lighting system — though we’ll have to wait and see how the new Philips Hue module handles third-party connectivity once it’s launched.

Despite these gray areas, the Nanoleaf-Hue integration is an unexpected yet possibly beneficial upgrade. When it comes to smart lighting brands, their respective devices tend to keep you locked into one ecosystem or the other, so having a simple but effective connectivity tool would give users much more flexibility.

Logitech is behind some of our favorite productivity peripherals and desk accessories - and right now, the company is running a big office upgrade sale as part of Amazon for Business Brand Week.

Two accessories catching my eye, that pair perfectly, are the Logitech MX Ergo S Advanced Wireless Trackball Mouse for $102 (was $120) and the Logitech Ergo K860 Wireless Ergonomic Split Keyboard for $128 (was $150) at Amazon.

Driving the deals for business users is Logitech's desire to highlight that 'bargain' office equipment comes with hidden costs for corporate buying teams.

Ergo K860 Wireless Ergonomic Split Keyboard: was $149.99 now $127.49
Curved split keyframe, integrated memory-foam wrist rest, adjustable palm lift, full-size layout, quiet scooped keys, Bluetooth and USB receiver connectivity, with customizable function keys.View Deal

MX Ergo S Advanced Wireless Trackball Mouse: was $119.99 now $101.99
Thumb-operated trackball, 20-degree adjustable tilt, six customizable buttons, Bluetooth, Logi Bolt receiver, USB-C charging, up to 120-day battery life and 80% quieter clicks.View Deal

Why we recommend them

The MX Ergo S mouse and Ergo K860 keyboard make sense together because they're both designed around making long periods at a computer more comfortable, but they approach the problem very differently.

The MX Ergo S keeps your hand largely stationary, with your thumb controlling the trackball instead of moving the mouse around your desk. Its 20-degree tilt, sculpted shape and quieter clicks make it an interesting alternative to a conventional mouse, while six programmable buttons add useful productivity options.

In his 4.5-star review of the older but very similar MX Ergo Wireless trackball, our expert Jon called it: "an exceptionally designed product that’s a joy to use. It might not have re-invented the wheel, but when it works this well it’s hard to ask for much more."

The Ergo K860 uses a curved, split design that places your hands at a more natural angle, backed by a substantial memory-foam wrist rest.

In her 4-star review, our expert Rosario said the "K860 keyboard does its job of providing support, comfort and reducing physical strain." She added it "has great features like its curved, split design, and a double-layered soft foam wrist rest with front end tilt legs. While initially awkward to use, you soon get used to it."

Both devices can connect wirelessly through Bluetooth or their included USB receivers, and Logitech's software provides additional customization.

Price context & historical value

The Logitech MX Ergo S is currently $101.99 at Amazon, down from $119.99, and the Ergo K860 is $127.49, down from $149.99, a 15% saving on both. We have seen the mouse cheaper in the past — it was briefly $84 earlier in the year — and the keyboard has been as low as $102. It's possible you might be able to pick up either or both devices for less in the future, but this is still a decent offer for office buying teams upgrading business gear for multiple staff.

Should you buy them?

✅ Buy the MX Ergo S and Ergo K860 if...

The Logitech MX Ergo S is worth considering if you spend long hours working and want to reduce how much you move your arm. Its thumb-controlled trackball, 20-degree tilt, six customizable buttons and up to 120-day battery life make it particularly well suited to productivity.

The Logitech Ergo K860 is good if typing comfort is your priority. Its curved split layout, pillowed wrist rest and adjustable palm lift are designed to encourage a more natural position.

❌ Skip the MX Ergo S and Ergo K860 if...

Give the MX Ergo S a miss if you've never got along with trackballs or prefer the speed and familiarity of moving a conventional mouse. It's also designed primarily for medium- to large-sized hands.

The Ergo K860 won't suit everyone either. Its split curved layout takes some adjustment. It's also a large keyboard at 17.95 x 9.17 inches so occupies a lot of desk space.

The Catch: What to know before you buy

It's worth noting that you need to be comfortable adapting to this keyboard and mouse set up. The MX Ergo S uses a thumb-operated trackball rather than traditional mouse movement and it's not one for left-handed users. The Ergo K860's curved split layout will initially also feel unusual if you've spent years typing on conventional keyboards.

• Shop all Logitech for Business deals Amazon