Feed aggregator

In states without policies to drive renewable energy, power prices could surge as federal tax incentives for clean energy disappear, according to Energy Innovation, a think tank.

(Image credit: Justin Sullivan)

The Trump administration is encouraging people to have more children, with baby bonuses and tax breaks. But some families who are practicing pronatalism want alternatives to hospital births.

(Image credit: Kendrick Brinson)

What truly separates us from machines? Free will, creativity and intelligence? But think about it. Our brains aren't singular, monolithic processors. The magic isn't in one “thinking part,” but rather in countless specialized agents—neurons—that synchronize perfectly.

Some neurons catalog facts, others process logic or govern emotion, still more retrieve memories, orchestrate movement, or interpret visual signals. Individually, they perform simple tasks, yet collectively, they produce the complexity we call human intelligence.

Now, imagine replicating this orchestration digitally. Traditional AI was always narrow: specialized, isolated bots designed to automate mundane tasks. But the new frontier is Agentic AI—systems built from specialized, autonomous agents that interact, reason and cooperate, mirroring the interplay within our brains.

Large language models (LLMs) form the linguistic neurons, extracting meaning and context. Specialized task agents execute distinct functions like retrieving data, analyzing trends and even predicting outcomes. Emotion-like agents gauge user sentiment, while decision-making agents synthesize inputs and execute actions.

The result is digital intelligence and agency. But do we need machines to mimic human intelligence and autonomy?

Every domain has a choke point—Agentic AI unblocks them all

Ask the hospital chief who’s trying to fill a growing roster of vacant roles. The World Health Organization predicts a global shortfall of 10 million healthcare workers by 2030. Doctors and nurses pull 16-hour shifts like it’s the norm. Claims processors grind through endless policy reviews, while lab technicians wade through a forest of paperwork before they can even test a single sample.

In a well-orchestrated Agentic AI world, these professionals get some relief. Claim-processing bots can read policies, assess coverage and even detect anomalies in minutes—tasks that would normally take hours of mind-numbing, error-prone work. Lab automation agents could receive patient data directly from electronic health records, run initial tests and auto-generate reports, freeing up technicians for the more delicate tasks that truly need human skill.

The same dynamic plays out across industries. Take banking, where anti-money laundering (AML) and know-your-customer (KYC) processes remain the biggest administrative headaches. Corporate KYC demands endless verification steps, complex cross-checks, and reams of paperwork. An agentic system can orchestrate real-time data retrieval, conduct nuanced risk analysis and streamline compliance so that staff can focus on actual client relationships rather than wrestling with forms.

Insurance claims, telecom contract reviews, logistics scheduling—the list is endless. Each domain has repetitive tasks that bog down talented people.

AI is the flashlight in a dark basement

Yes, agentic AI is the flashlight in a dark basement: shining a bright light on hidden inefficiencies, letting specialized agents tackle the grunt work in parallel, and giving teams the bandwidth to focus on strategy, innovation and building deeper connections with customers.

But the true power agentic AI lies in its ability to solve not just for efficiency or one department but to scale seamlessly across multiple functions—even multiple geographies. This is an improvement of 100x scale.

1. Scalability: Agentic AI is modular at its core, allowing you to start small—like a single FAQ chatbot—then seamlessly expand. Need real-time order tracking or predictive analytics later? Add an agent without disrupting the rest. Each agent handles a specific slice of work, cutting development overhead and letting you deploy new capabilities without ripping apart your existing setup.

2. Anti-fragility: In a multi-agent system, one glitch won’t topple everything. If a diagnostic agent in healthcare goes offline, other agents—like patient records or scheduling—keep working. Failures stay contained within their respective agents, ensuring continuous service. That means your entire platform won’t crash because one piece needs a fix or an upgrade.

3. Adaptability: When regulations or consumer expectations shift, you can modify or replace individual agents—like a compliance bot—without forcing a system-wide overhaul. This piecemeal approach is akin to upgrading an app on your phone rather than reinstalling the entire operating system. The result? A future-proof framework that evolves alongside your business, eliminating massive downtimes or risky reboots.

You can’t predict the next AI craze, but you can be ready for it

Generative AI was the breakout star a couple of years ago; agentic AI is grabbing the spotlight now. Tomorrow, something else will emerge—because innovation never rests. How then, do we future-proof our architecture so each wave of new technology doesn’t trigger an IT apocalypse? According to a recent Forrester study, 70% of leaders who invested over 100 million dollars in digital initiatives credit one strategy for success: a platform approach.

Instead of ripping out and replacing old infrastructure each time a new AI paradigm hits, a platform integrates these emerging capabilities as specialized building blocks. When agentic AI arrives, you don’t toss your entire stack—you simply plug in the latest agent modules. This approach means fewer project overruns, quicker deployments, and more consistent outcomes.

Even better, a robust platform offers end-to-end visibility into each agent’s actions—so you can optimize costs and keep a tighter grip on compute usage. Low-code/no-code interfaces also lower the entry barrier for business users to create and deploy agents, while prebuilt tool and agent libraries accelerate cross-functional workflows, whether in HR, marketing, or any other department.

Platforms that support PolyAI architectures and a variety of orchestration frameworks allow you to swap different models, manage prompts and layer new capabilities without rewriting everything from scratch. Being cloud-agnostic, they also eliminate vendor lock-in, letting you tap the best AI services from any provider. In essence, a platform-based approach is your key to orchestrating multi-agent reasoning at scale—without drowning in technical debt or losing agility.

So, what are the core elements of this platform approach?

1. Data: Plugged into a common layer

Whether you’re implementing LLMs or agentic frameworks, your platform’s data layer remains the cornerstone. If it’s unified, each new AI agent can tap into a curated knowledge base without messy retrofitting.

2. Models: Swappable brains

A flexible platform lets you pick specialized models for each use case—financial risk analysis, customer service, healthcare diagnoses—then updates or replaces them without nuking everything else.

3. Agents: Modular workflows

Agents thrive as independent yet orchestrated mini-services. If you need a new marketing agent or a compliance agent, you spin it up alongside existing ones, leaving the rest of the system stable.

4. Governance: Guardrails at scale

When your governance structure is baked into the platform—covering bias checks, audit trails, and regulatory compliance—you remain proactive, not reactive, regardless of which AI “new kid on the block” you adopt next.

A platform approach is your strategic hedge against technology’s ceaseless evolution—ensuring that no matter which AI trend takes center stage, you’re ready to integrate, iterate, and innovate.

Start small and orchestrate your way up

Agentic AI isn’t entirely new—Tesla’s self-driving cars employs multiple autonomous modules. The difference is that new orchestration frameworks make such multi-agent intelligence widely accessible. No longer confined to specialized hardware or industries, Agentic AI can now be applied to everything from finance to healthcare, fueling renewed mainstream interest and momentum. Design for platform-based readiness.

Start with a single agent addressing a concrete pain point and expand iteratively. Treat data as a strategic asset, select your models methodically, and bake in transparent governance. That way, each new AI wave integrates seamlessly into your existing infrastructure—boosting agility without constant overhauls.

We list the best IT Automation software.

This article was produced as part of TechRadarPro's Expert Insights channel where we 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/news/submit-your-story-to-techradar-pro

An "MP3" or digital audio player can help you take the tunes offline or at the very least off your phone.

Europe has begun to actively test the waters of cloud sovereignty, with 72% of European businesses prioritizing data control when selecting technology vendors. However, despite the growing desire to protect data integrity and security within European borders, over 70% of European businesses use US hyperscalers from their cloud computing provider.

While Google is doubling down on its commitment to data sovereignty, there is a growing concern over hyperscalers' dominance over the European market, as it leaves critical infrastructure in the hands of dominant foreign providers.

As sweeping tariffs continue to escalate tensions between Europe and the US's big tech, many are questioning whether Google's commitment is enough to protect Europe’s data from the Big Three.

US influence on Europe’s tech ecosystem

US Policies, like the 2018 Clarifying Lawful Overseas Use of Data (CLOUD) Act, give US hyperscalers massive influence in Europe. The act grants US authorities and federal agencies access to data stored by US cloud service providers, even when hosted in Europe. This raises concerns about whether European data stored with US-based providers can ever truly be sovereign, even if housed within European borders.

Another concern in Europe: being cut off from US services. If Europe were to suddenly lose access to US cloud services or face rising costs, businesses would experience immediate setbacks, from service disruptions to escalating operational expenses. These concerns, along with a push for more leadership, independence, and economic competitiveness, have led to Europe steadily building its own cloud ecosystem – fostering a network of regional providers and implementing policies that reinforce data independence.

The question has now become, do these changes signal a true tipping point for Europe? Or are they merely the first steps in a much longer transformation?

Who is driving the adoption of sovereignty?

Both the public and private sectors play pivotal roles in advancing cloud sovereignty across Europe. Governments have established regulatory frameworks to enhance standards and avoid fragmentation. However, policymaking is often slow and subject to political negotiation, making private sector initiatives crucial in accelerating the shift toward true sovereignty.

The private sector has emerged as a driving force behind the practical implementation of sovereignty ideals. According to Dominique Tessier, Head of Cybersecurity Focus Group at the European Champions Alliance (ECA), “the move to make sure the EU Cloud Certification Scheme will finally include an “upper security layer” is mainly driven by private European companies, as AIRBUS, EDF, Telecom Italia and others, whose efforts are gaining momentum”.

While companies like Microsoft have invested heavily in EU infrastructure to comply with local regulations, concerns remain about whether this represents true sovereignty or just a regulatory workaround. In contrast, European companies and partnerships, such as the joint venture between OVHcloud and Capgemini, are working to offer services independent of US control, aiming to create fully sovereign cloud solutions.

These initiatives reflect a growing acknowledgement of the strategic importance of cloud sovereignty. This is supported by Rahiel Nasir, Research Director, IDC Europe, who states that “interest in sovereignty has moved from governments and regulated sectors to all industry sectors, especially in Europe, and everywhere else where cloud is just beginning to pick up”. These efforts are becoming more widespread, indicating a collaborative push towards achieving European digital independence, but more needs to be done to make this achievable.

How can Europe achieve ‘true sovereignty?’

Achieving true cloud sovereignty requires more than just localized data storage, it demands complete independence from hyperscalers. To achieve this, Europe must go beyond compliance and foster a robust ecosystem of local providers that can match and work alongside hyperscalers.

While hyperscalers play a role in the broader cloud landscape, they should not be relied upon for sovereign data. According to Tessier, “the new US Administration has shown that it won’t hesitate to resort either to sudden price increases or even to stiffening delivery policy. It’s time to reduce our dependencies, not to consider that there is no alternative”.

For Nasir, the key is striking a balance. “In an ideal scenario, local providers and global providers should partner for sovereignty to work at scale”. Leveraging their capabilities where appropriate while ensuring critical data and workloads remain within truly sovereign infrastructure.

By shifting away from hyperscaler dependency and building a diverse, sovereign infrastructure, organizations can move beyond regulatory compliance and achieve operational independence within their own jurisdictions.

The path to sovereignty

While Europe is leading the way in advocating for cloud and digital sovereignty, achieving true independence requires a strategic balance. Reducing reliance on US hyperscalers while developing competitive local alternatives is crucial. This balance involves leveraging public and private sector initiatives to create an environment where local providers can thrive and compete on a global scale.

Ultimately, sovereignty is not just about regulatory compliance; it's about a strategic vision for independence. Empowering local providers and creating interconnected networks will allow Europe to set its own digital agenda and drive long-term economic and technological growth, helping to achieve “true” sovereignty.

We list the best cloud storage.

This article was produced as part of TechRadarPro's Expert Insights channel where we 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/news/submit-your-story-to-techradar-pro

We consulted dentists to discover the potential benefits and side effects of fluoride-free, nano-hydroxyapatite toothpaste.

AI is democratizing access to software development in new and innovative ways, with 'vibe coding' emerging as the latest buzzword for budding developers. For the uninitiated, vibe coding makes it easy for anyone to get a head start on projects, by merely describing what they want AI tools such as Cursor, GitHub’s Copilot or Replit to accomplish.

This is no small thing, especially for those who before had not been able to create software on their own. However, vibe coding requires a high level of trust in the AI’s output, and there are potential trade-offs in confidence and security for a faster turnaround and expanding capabilities to those who would not otherwise be able to code.

Exploring the latest AI technologies

Exploring the latest AI technologies can undoubtedly help developers experiment with new ways in which to do things better and faster, and vibe coding is no exception. However, even Andrej Karpathy, the former AI director at Tesla who coined the term, advises that the methodology is better suited for “weekend projects” than for more complex projects.

For smaller, proof-of-concept style projects, the speed of vibe coding can shine, but as things grow so does the need for important context and knowledge of edge cases. Vibe coding practicalities aside - recognising security exploits, lack of contextual understanding, bug fixing and software life cycle/maintenance considerations – there’s also a larger issue at play.

For AI-powered coding grunt work to improve in any format or scale, socially responsible AI must be the foundation on which technology solutions are built and delivered. The more trusted and accurate the data that large language models train on, the higher quality the outcome - for the code and most importantly, for the larger tech community.

There is a need for a symbiotic relationship to form: data helps create and improve AI experiences, and AI experiences help guide new, human verified information.

Understanding the opportunities and risks associated with vibe coding

AI continues to democratize access to software development in new and innovative ways for aspiring developers. Developers can ask AI tools such as Cursor, GitHub’s Copilot or Replit to help initially scaffold what they would like to see at the beginning of a new project. This is where AI generated code can be very powerful, and I use it on my own homegrown projects.

There is little need for context, edge cases are much fewer, and security concerns are more “standard” than other larger projects with bespoke needs. However, the trade-off for improved speed can be a sacrifice in potential security concerns as those code bases grow.

For budding programmers, vibe coding has the potential to provide the necessary support during the early stages of a project, and anything that gets more people into our field and shortens the learning curve to coding is always a good thing. But caution should also be exercised given the risks associated with the method for things outside of some early, base use cases.

Handing over the reins to AI

Andrej Karpathy describes vibe coding as interacting with AI to assist developers shift away from manual programming activities through LLMs (large language models), prompting developers through intuitive decisions to make it easier to create software as LLMs continue to improve code writing skills.

To fully embrace vibe coding, developers must cede much of the control to the AI assistant during the entire process, rather than becoming aware and having an understanding of the code as it is put into the codebase.

As LLM assistants continue to improve their corpus of knowledge on developing code, real-time decisions and predictions are made around what you would like to happen next to successfully complete a project, but these are still assumptions and educated guesses based on the experiences of others.

Vibe code with caution

It should always be remembered that every coding process cannot be overseen exclusively with the help of AI assistance. Small, low-risk side projects are ideal for vibe coding. When it comes to larger, more complex projects – a human should always be a first-class part of that loop.

AI coding tools powered by LLMs can and do produce mistakes. Developing larger datasets of information and considering other factors such as quality control and security requires an expert eye to monitor flaws or weaknesses that could be new or unexpected to an AI who is only thinking of other use cases, or the most common ones.

Knowledgeable developers are able to identify and test vulnerable code themselves – LLMs are sometimes simply unable to even register any mistakes they may produce.

In short, what the AI assistant doesn’t know, it will not flag as an issue – be that a bug or security vulnerability, AI can be quite confidently wrong. Vibe coding is a new trend for the industry but cannot be seen as a silver bullet or license for cutting corners in the development process when quality and stability matter.

Human expertise must always contribute to the process along the way as code bases grow, either vertically or horizontally.

We list the best site for hiring developers.

This article was produced as part of TechRadarPro's Expert Insights channel where we 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/news/submit-your-story-to-techradar-pro

An appeals court late Monday stepped in to keep in place protections for nearly 12,000 Afghans that have allowed them to work in the U.S. and be protected from deportation.

(Image credit: Omid Haqjoo)

Here are the answers for The New York Times Mini Crossword for July 15.

• Claude can now connect to apps like Notion, Canva, and Stripe
• The AI can understand and assist with tasks using your actual work data
• Claude's secure access reduces the need to constantly explain context to the AI

Anthropic has upgraded Claude with a major new set of tools that let the AI assistant integrate directly with several popular software tools, including Notion, Canva, Stripe, Figma, Socket, and Prisma. The new Claude tool directory means you don't need to explain what you want to Claude every time you want to employ those tools; Claude can now look at the same information as you to help.

Until now, most AI interactions have required copying and pasting every detail from your project management tool, explaining what’s important, clarifying what each task means, and double-checking that the AI understood it. Now you can just ask it to do the task, and Claude will pull the information directly from the relevant tool to handle things.

That might not seem groundbreaking at first glance, but that context gap is where things usually fall apart when asking AI chatbots to help you. For instance, if you're working on a product launch in Notion and have a list of things to do, you'd normally have to retype or upload all the information to Claude. Now, once you connect Notion to Claude, the AI can read your project documents directly and start putting together timelines and presentation materials that fit the product because it sees what you see.

Or imagine a small business owner using Stripe to manage payments who wants a summary of which customers paid last week and which still owe you for your services. Claude can now pull that data directly from Stripe with your permission. And with Canva, a blank social media post template can now be filled in with a design and copy from Claude based on your brief. You describe what you need in plain language, and Claude will make something usable.

Claude connected

These integrations are powered by something called the Model Context Protocol, or MCP. That basically means Claude can understand and act on the tools you use without needing a whole tutorial. You just connect an app once, and Claude gets secure, limited access to the relevant information inside it. It doesn’t read your entire inbox or download your bank history, just what’s necessary to help you with the task at hand.

You can go to Claude’s tool directory and connect whatever apps you already use. If you’re on a paid Claude plan, you’ll get access to remote app connections like Stripe and Notion. Desktop integrations, like Figma and Socket, are available through the Claude desktop app.

Other AI tools are trying something similar. Google’s Gemini shows up in Docs and Gmail. Microsoft’s Copilot is baked into Word and Excel. But Anthropic’s take is more about linking what you already do with the AI, as opposed to baking the AI into those apps directly.

Of course, this doesn't make Claude autonomous. It can’t pay your bills or fully run your job. And while Anthropic says it’s designed everything with privacy and security in mind, some are likely to be wary, even if you can choose what Claude can access. But for most regular users, this update represents something potentially very useful in staying on top of things. If, as Anthropic claims, it will save time and mean you don't have to redo a lot of tedious paperwork, it will likely be a very popular feature.

You might also like

• How Claude’s 3.7's new ‘extended' thinking compares to ChatGPT o1's reasoning
• I tried Claude's new Research feature, and it's just as good as ChatGPT and Google Gemini's Deep Research features
• Claude goes to college and wants to be your study buddy

Narwal Freo Z10: Two-minute review

I’ve tested quite a few brands of robot vacuums in the last decade, but this is my first Narwal. I was quite excited to try it out as I’ve read good things about the company’s previous robot vacuums from TechRadar’s other reviewers. That includes our Narwal Freo Z Ultra review, where it “smashed [its] mopping tests”.

The Narwal Freo Z10, a slightly cheaper version of the Freo Z Ultra, is very much cut from the same cloth – it too is a fantastic mop but, like its more expensive sibling, carpets are not its forte. Actually, I will go so far as to say that vacuuming in general isn’t its forte.

It’s 15,000 Pa suction looks good on paper, but it took two consecutive vacuum runs at its highest suction setting for it to pass my vacuuming tests. And that’s just hard floors. It struggles even more on carpets, although if you are patient, it will clean about 80% of hair off the fibers, but that’s not a lot considering there are other robot vacuums I’ve tested recently that do a better job.

The other big issue with the vacuuming is the self-cleaning of the onboard dustbin. Even though I can hear it say “self-cleaning started” and it makes the appropriate noises, dust and debris from several vacuuming sessions just kept compacting inside the onboard bin to a point where I had to pull out the hair and dirt myself to increase suction.

And while its anti-tangle brushes do what they promise, I found that the hair strands it invariably misses sucking up from hard floors get collected by the mop pads as clumps that I have to pick up myself.

So while I question the Freo Z10’s vacuuming prowess, I can’t fault its mopping – and the evidence is there to see in the color of the dirty mop water. The kind of tiles I have in my apartment require high waterflow rates, and the Narwal Freo Z10’s Wet Mopping setting was perfect – however, this mode guzzles water. I had to refill the clean water every time I did a full-home cleaning. If you have wood floors, though, you shouldn’t need to worry as the lower water settings will treat the planks kindly.

While I didn’t mind refilling fresh water in the middle of a cleaning session, I can’t quite forgive how basic the map in the app is. It can be edited, of course, but that’s fiddly to do if your rooms aren’t precisely placed squares or rectangles.

Even that is a small issue compared to my experience with single-room cleaning. This defaults to the Customized Mode option in the app and I think it's badly named. The assumption would be that you can set up custom cleans here (those are, in fact, saved as Shortcuts, by the way), but you have to keep editing it for every room clean you want to do – your living room may not need as much cleaning as your dining room, for example, and switching from one to the other means you’re constantly editing that mode.

Then there’s the Freo Mind automatic cleaning mode which, in theory, uses AI to determine the best clean for the room or full house, but oh boy, does it get it wrong most of the time. During my testing, it invariably cleaned my living room multiple times before moving on to do just a basic cleaning in the rest of the apartment. In one such session, it mopped the living room four times, the dining room twice, while the entry hallway and the home office got just a single mop!

The one really good thing about the Freo Z10 is its intelligent use of detergent. It automatically takes some only when it thinks a floor space needs deeper mopping. This automatic detergent use happens in the Freo Mind mode and I could tell when it was using the soap by keeping an eye on the dirty-water tank for suds after each room’s mopping was done over multiple sessions.

Despite its excellent mopping capabilities, it’s hard to get over the fact that the onboard dustbin doesn’t self-empty properly and the app isn’t well thought out either. For just a robot mop, it’s priced too high.

(Image credit: Sharmishta Sarkar / TechRadar)Narwal Freo Z10 review: Price & availability

• First shown at CES in January 2025; released April 2025
• Available in the US and Australia, but not in the UK
• List price: $1,099 / AU$1,999

Despite being cheaper than the flagship Narwal Freo Z Ultra (which retails for $1,499/ AU$2,499), I can’t really say that the Freo Z10 is ‘affordable’ at its full price of $1,099 / AU$1,999. I can’t even say it’s competitively priced because it’s not quite the efficient cleaner I was hoping it would be.

The only other robot vacuum cleaner that comes in at this price point that I would recommend is the Ecovacs Deebot T30 Omni (£799 / AU$1,749), but it isn’t widely available and it's got no detergent dispenser. This cheaper Deebot even beats out the Freo Z10 in terms of performance too, vacuuming and mopping well, although the Narwal is the better mop despite boasting higher suction (15,000 Pa vs 11,000 Pa).

If you want more intelligent features, which the Freo Z Ultra can provide, you’re obviously going to need to spend more. For example, the Dreame L40 Ultra available in the US and the UK with 11,000Pa suction has a list price of $1,499.99 / £999 (unavailable in Australia), but take it up a notch and the 12,000Pa Dreame X40 Ultra with detachable mop pads will set you back a whopping $1,899 / £1,299 / AU$2,799. Similarly, the Eufy Omni S1 Pro, with its interesting slimline design but with a maximum of 8,000Pa of suction costs $1,499 / £1,499 / AU$2,699.

Of course, as with any all-in-one robot vacuum, you will also need to take into consideration the ongoing cost of maintaining the Freo Z10. Replacement dust bags will cost you $19.99 in the US for a set of two, while it’s AU$29.99 for a set of three Down Under. Similarly, when it comes to replacing the mop pads, a pair will cost you $19.99 / AU$39.95. That said, each dust bag will last you up to 120 days and the mop pads a lot longer.

What you might require more frequently will be the detergent, which will set you back $29.99 / AU$59.95 for the 19.6oz/580ml pack. Again, I say ‘might’ need frequently – this will depend on how often the machine thinks it needs to use soap on your hard floors. So that 19.6oz/580ml bottle might well last months.

• Value score: 3.5 / 5

(Image credit: Sharmishta Sarkar / TechRadar)Narwal Freo Z10 review: specs

Suction power:

15,000Pa

Onboard bin capacity:

Up to 1L (compressed)

Dust bag capacity:

2.5L

Clean water tank capacity:

4.5L

Dirty water tank capacity:

4L

Reservoir capacity:

N/A

Mop lift:

12mm

Noise level:

up to 55dB

Battery capacity:

5,200mAh

Runtime:

up to 210 minutes

Charging time:

up to 2.5 hours

Dimensions (L x W x H):

Robot = 355 x 350 x 107.7 mm; base station = 431 x 426.8 x 461.5 mm

Weight:

Robot = 4.6kg; base station = 11.3kg

Narwal Freo Z10 review: Design

• Curvy but solid base station
• Robot tucks right inside base station
• Features a detergent dispenser

The Freo Z10 robot vacuum isn’t very different from other vacuum bots, featuring a small turret on top to house some of its sensors, and it's quite similar to the one supplied with the Narwal Freo Z Ultra. The main difference between the two siblings is that the Freo Z10 lacks the camera that the Ultra model has.

I do love the curvaceous look of the base station, though, which is unique to the Freo line. It’s quite different.

What I didn’t expect is how heavy the whole thing would be to take out of the box it came in. You see, the robot is already tucked inside the base station in the packaging, so you’re essentially trying to lift about 11 kilos in one go out of the box. Other robot vacuums come with various parts packed separately, making them easier to set up out of the box.

Well, if you’re prepared to haul out a heavy machine, the rest of the setup is simple. You just need to fit the cleaning tray inside the base station, slip the robot in, then attach the base station extension ramp.

Compared to some other robot vacuums I’ve tested recently, the Freo Z10's base station is broad, so be sure you have enough space around it for the robot to safely maneuver in and out of the base station.

Image 1 of 4

(Image credit: Sharmishta Sarkar / TechRadar)Image 2 of 4

(Image credit: Sharmishta Sarkar / TechRadar)Image 3 of 4

(Image credit: Sharmishta Sarkar / TechRadar)Image 4 of 4

(Image credit: Sharmishta Sarkar / TechRadar)

What sets the Freo Z10 apart from many other brands is that it has two side brushes rather than just one. Both feature two tufts of bristles, but one has them placed closer together, while the other is further apart.

The side brushes turn at different speeds sometimes, depending on what the Freo Z10's DirtSense technology is telling it, or if it’s along a room edge. On paper, the varying speeds should help it push hair and dirt more effectively towards the central bar brush under the robot, but in reality, that doesn’t work as well as I had hoped (more on this in the Performance section).

The central bar brush is anti-tangle and, in the six weeks I’ve used the Freo Z10, I’ve never had to clean it. It pulls everything into a small onboard dustbin that, unlike other such robovacs, doesn’t have a lid, but the filter can be taken off for cleaning.

While it’s supposed to self-empty each time it finishes a vacuum run, I found that dirt just kept compacting inside and I had to manually pull it all out via the two holes (one facing the bar brush, the other towards the base station) as it was beginning to affect the suction.

The mop pads are slightly triangular in shape to help them get into corners, which they do better than most of the rounded mop pads on other similar robovacs.

Image 1 of 2

(Image credit: Sharmishta Sarkar / TechRadar)Image 2 of 2

(Image credit: Sharmishta Sarkar / TechRadar)

The curvaceous base station features a deep cavity that practically engulfs the entire robot when it’s docked, something I haven't seen on any other brand of robot vacuum I've tested. I like that the bot is barely visible, it's a clean aesthetic.

As with similar models, the base station has the capability of emptying out the onboard dustbin into a 2.5L dust bag which, incidentally, is supposed to compact the dirt to accommodate about 3L. As I’ve mentioned though, the self-empty did not work for me and the dust bag remains nearly empty even after six weeks of use at the time of writing.

Beside the cavity for the dust bag is space to fit a bottle of detergent and one comes in the box – the Freo Z10 will automatically take detergent only when it’s needed, which is an intelligent function I appreciate.

On the top of the base station are a few buttons that allow you to control the robot without needing to reach for your phone each time. And under the top cover are two plastic tanks – one for clean water and the other for the dirty mop water.

With some silver accents on the white unit, the Freo Z10 is a well-designed robot vacuum – I just wish it performed as good as it looks.

• Design score: 4 / 5

(Image credit: Sharmishta Sarkar / TechRadar)Narwal Freo Z10 review: Performance

• Excellent mopping and good obstacle avoidance
• High suction on paper doesn’t translate in performance
• Auto-emptying of dustbin doesn’t work; mops get washed even when not used

Starting off with using the Narwal Freo Z10 was a wonderful experience – it connected to my home’s Wi-Fi instantly and was set up within a minute, ready to map my apartment.

Mapping took about six minutes for the one-bedroom space, but I have to admit I wasn’t quite enamored with the final map that was created (more on this later in the App Control section).

The test space for this review, which measures approximately 40sqm, includes wall-to-wall carpet in the bedroom, with the rest being tiled floors throughout.

The Freo Z10’s first cleaning run was fantastic! It did a decent enough job on the carpet (although I have tested others that did slightly better), but I was thrilled with how well it mopped.

Vacuuming

My initial vacuum run with the Freo Z10 was satisfactory, so I went into full testing mode with high expectations from its 15,000 Pa of suction.

It has four vacuum settings – Quiet, Normal, Strong and Super Powerful – and you can choose to have it vacuum up to three times in a single session. One thing I have to note right up front: even on the Strong setting, the Freo Z10 is comparatively very quiet. It’s only when you switch to Super Powerful that it truly makes its presence felt.

In my testing, I wasn’t particularly impressed with the Quiet and Normal settings as it barely sucked up anything – hair strands and visible dust remained on the hard floors and carpets looked untouched. Strong, though, is fine for hard floors, although I found it needed to be set to Super Powerful for the best possible vacuum on carpets.

Case in point: while performing TechRadar’s standard tests of vacuuming oats and tea on carpets and hard floors, I found that the Freo Z10 needed two sessions in the highest suction setting to pick up all the oats from tiles, but I could still see some tea on the tiles (to be fair it’s tea from a bag, so fine particles). I also sprinkled some talcum powder to test suction of very fine particles and it wasn’t able to do a good job even after two passes. I had to use my Dyson V15s Detect Submarine to clean the remaining tea and talcum powder off the tiles.

It was a similar experience on carpets, although here it even struggled to clean up the larger oats pieces too. In the bedroom, where my hair collects on the carpet fibers each day, it managed to clean about 80% of the strands, so a little disappointing compared to the Ecovacs Deebot T30 Omni and the Roborock Qrevo Edge I've previously tested in the same area.

Long story short, I’m really not impressed with the Freo Z10’s vacuuming prowess, but it works in a pinch if you need a quick job done before guests arrive – provided the onboard dustbin isn't already full to reduce overall suction.

Mopping

Interestingly the Freo Z10 robot doesn’t have a water reservoir onboard – it just goes back to the doc to repeatedly wet its mop pads. Despite that I was quite impressed with how well it mopped – a pleasant relief after seeing its lack of vacuuming strength.

Its slightly triangular mop pads can manage to get into corners, but temper your expectations for corner cleaning as there's just not enough pressure exerted on the edges of the mop pads to be effective. On the rest of the floor though, the 8 Newtons of pressure it exerts can mop up dried and wet stains well, although the waterflow rate you choose will determine how well.

On the kind of tiles I have in my apartment, the Wet Mopping option was best for dried-up dusty stains and wet sauce spills. A single run was good enough to clean up the former but I found it sometimes needed to be run twice in an area to clean up the latter. It's a slightly different story if there's dried sauce spills on the floor – three turns in the Wet Mopping mode did the trick during my testing, although it will depend on what exactly it's cleaning up.

(Image credit: Sharmishta Sarkar / TechRadar)

For everyday mopping, however, I found the Normal waterflow option to work well and, if you have wood floors, the Low setting will treat them kindly.

One thing to note though: the Freo Z10 guzzles water much more than other robot vacuums I've tested, particularly on the Wet Mopping mode, and I had to refill the clean-water tank every time it did a full-house mop. It's not a deal breaker because it really does a good job.

Another thing to keep in mind about its mopping is related to its lack of vacuuming: because it doesn't always pick up all the hair or fur from a floor, I found the mop pads gather the strands and leave them on the floor as damp clumps. Not ideal at all and that means it's hard to recommend it for households with pets.

Navigation and obstacle avoidance

Complementing the mopping is the Freo Z10's obstacle avoidance. It's the first robot vacuum that I've tested that was able to avoid cables in its path. While it tried to swallow up a very slim Lightning cable for an iPhone, it easily avoided three other USB-C cables I put in its path while testing.

Anything a little bigger and that's an even easier job for the Freo Z10 to avoid. Footwear, slim furniture legs, even little tea-light holders (I have a floor-height shelf under my entertainment unit that most robovacs climb onto) are all left in peace. However, beware the sock on the floor – the Freo Z10 will try and swallow little booties or no-show socks, like practically every other robot vacuum.

Its navigation is a mixed bag, though. It's fine as long you don't move things around after its map has been created, but shift a chair even a little bit and it does tend to get confused.

I noticed this first when I left a large package in a corner that was cleaned on a previous run. While it avoided banging into the box, it moved around it a few times to figure out it couldn't get to that space it had previously roamed. To test this again, I moved my living room chairs a little (just by a couple of inches) and, lo and behold, the precise back and forth movement the Freo Z10 usually employs was messed up and that also meant it went around in circles a few times before getting back to its dock after its job was done.

It never got lost during my testing, but it definitely doesn't quite like things out of place.

Base station performance

Like many other similar robovacs, the Freo Z10's base station is multipurpose, requiring minimal human intervention and, for the most part, that's true. All you need to do is remember to refill the clean water as needed and empty out the dirty so it doesn't begin to grow its own little ecosystem.

It's mop washing is arguably the best I've seen in any robot vacuum. Even after six weeks of use, the mop pads still look quite pristine despite the color of the dirty water being, well, very dirty. I was impressed with the Deebot T30 Omni's washing, but this surpasses even that. Narwal's use of hot-water washing and hot-air drying is very effective indeed.

This could perhaps be attributed to the fact that the mops get washed even if they didn't get used. Every time I did a vacuum-only run (whether it was a single room or the full house), the mops always got washed. I hadn't experienced this with the previous robovacs I'd tested, so it took me by surprise. This also meant that the base station was consuming more power than was necessary as the hot-air drying can take a few hours.

Image 1 of 2

(Image credit: Sharmishta Sarkar / TechRadar)Image 2 of 2

(Image credit: Sharmishta Sarkar / TechRadar)

Another feature I appreciate here is the 'disinfecting' of the bin and dust bag. It uses hot air to keep things dry so as to avoid growing bacteria and fungus inside and, after weeks of use, there's no smell anywhere to indicate otherwise.

Now, I can forgive the excessive mop washing I suppose but, as I've already mentioned, the lack of self-emptying of the robot's dustbin is worrying. Where most robot vacuums return to their docks to repeatedly empty their bins in the middle of a vacuum run, the Freo Z10 doesn't – it waits till the job is finished to do so. I can hear the robovac say "self-cleaning started" and the base station suction can be heard but, obviously, no emptying takes place and this subsequently starts affecting the suction power of the bot.

Maintenance of the unit involves needing to refill and clean out the tanks and, every few days, washing and drying the filter on the bin of the bot. What should be minimal intervention, however, also included emptying the bin myself after a few uses, which is far from ideal.

(Image credit: Sharmishta Sarkar / TechRadar)Battery life

Narwal says that the Freo Z10's 5,200mAh battery will run up to 210 minutes on a single charge, but that of course depends on what settings you use.

Despite having it set to max suction (Super Powerful in the app) and at the highest waterflow rate, I found the Freo Z10 could do a single full-house vacuum-then-mop session without needing to go back to the mains to charge.

It needed to top up when it was set to 2x vacuum and 2x mop in the Meticulous cleaning pattern.

The Freo Z10 runs all the way down to about 5% before making its way back to the base station and, in my case, juices up till it's about 45% to complete a job if may have left unfinished.

So when it comes to battery life, I really can't complain.

• Performance score: 3 / 5

Narwal Freo Z10 review: app control

• Basic map that's fiddly to edit
• Able to save plenty of Shortcut routines
• Freo Mind AI mode is not efficient

If there's one thing all robot vacuums have in common is a subpar app experience. I've said this before with the Ecovacs app and a colleague has complained about it with Dreame too. The only brand that has a wonderful, albeit over-the-top, app experience is Roborock.

Sadly the Narwal Freo app can't compete with Roborock and is even more disappointing than the Ecovacs Home app.

Now, don't get me wrong, it's perfectly functional, but the map is very basic and if your home isn't divided into near-perfect squares and rectangles that align well, it's hard to edit the map to accommodate rooms perfectly. In my case, I couldn't edit the entry hallway in my apartment to the correct size, which I was able to do in the Roborock and Ecovacs apps.

That's not a deal breaker, to be honest, and the Narwal Freo app actually does one thing better than Ecovacs – offer a Shortcuts list that accessible from the device's home page. You can set up several routines in this list and it makes for quick and easy access for frequently used cleaning runs (for example, carpet-cleaning only or after-dinner cleaning).

(Image credit: Narwal)

What's confusing is a Customized Mode within the app that it defaults to when you pick a room or zone cleaning. While I can see why it's there, it's rather annoying as you constantly have to edit that one option for each room or zone clean.

Case in point: I clean my dining room more often than the rest of the apartment but each run has different requirements. I might have set it to do a vacuum-then-mop session one afternoon, but the next time I might want just mopping and I have to edit the Customized Mode for each. The next day I might need to vacuum my bedroom only (which is carpeted) and I have to delve into the Mode to change things.

It doesn't take long, but when using it every day or two, it does become rather annoying. Other brands have a more seamless room or zone cleaning app experience in my opinion.

Image 1 of 2

(Image credit: Narwal)Image 2 of 2

(Image credit: Narwal)

Like other premium robot vacuums released in the last year or two, the Freo Z10 also features an AI mode called Freo Mind. In theory, this automatically picks the best cleaning option for your home based on previous sessions but, in reality, I found it to be very inefficient.

In the six weeks of testing, it has cleaned my dining room more than any other space in the apartment, but it doesn't leverage that knowledge. Instead, it chooses to do multiple cleaning sessions in my living room, which hardly ever gets dirty (the dining room is where everyone congregates), but does maybe one or two cleans elsewhere.

This means it's using up its battery life, so chances are it will take the entire day to finish cleaning a home when set to Freo Mind as it will need to repeatedly charge itself. It will also require you to refill the clean water repeatedly if it mops one space multiple times.

After trying out Freo Mind a few times, I gave up on it and never went back – Narwal will need to work on it to make it a little more efficient.

• App score: 3.5 / 5

(Image credit: Sharmishta Sarkar / TechRadar)Should I buy the Narwal Freo Z10?

Attribute

Notes

Score

Value

It's not cheap, but compared to similar options, it's well priced. However, taking its overall performance into account, it's an expensive mop.

3.5 / 5

Design

It's a well-designed, sturdy robot vacuum with a curvaceous base station that I'm personally fond of.

4 / 5

Performance

While its mopping and obstacle avoidance are good, the basic task of vacuuming isn't its best attribute. Moreover, the lack of self-emptying is concerning.

3 / 5

App control

With a fiddly map, an annoying Custom mode and subpar Freo Mind performance, the Narwal Freo app needs work.

3.5 / 5

Buy it if...

Mopping is important to you

If there's anything the Narwal Freo Z10 does better than many other robot vacuums, it's mop. Moreover, its mop washing is one of the best I've seen.

You have a lot of things lying around the house

Home owners with clutter on the floor will appreciate the Freo Z10's obstacle avoidance. Sure, it might swallow up a tiny Lego piece, but it's one of the few robovacs I've tested that doesn't get tangled up in cables.

Don't buy it if...

You want good vacuuming prowess

Mopping might be the Freo Z10's forte, but the basic task of vacuuming is questionable, especially since I found the self-emptying isn't what it's advertised to be. If you have pets, you'll want to look elsewhere for a better robot vacuum.

You want an efficient robot vacuum

The whole idea behind a premium robovac is automated cleaning, but Narwal's Freo Mind isn't the best option to use, so you will need to set up several routines. Moreover, constantly washing mops even when they haven't been used isn't the best consumption of power.

You want a seamless app experience

Most robovacs don't have a good app, but Narwal's Freo is arguably one of the most basic ones I've used.

Narwal Freo Z10 review: Alternatives

Ecovacs Deebot T30 Omni

If you can find the Ecovacs Deebot T30 Omni where you live, I think it's one of the best-value robot vacuums you can get, and it's often discounted too. It has all the essential features of a premium robovac, and an interesting design as well, and performs well for a bot that has a maximum of 11,000Pa of suction
Read my in-depth Ecovacs Deebot T30 Omni review

Narwal Freo Z Ultra

If it's a Narwal you have your heart set on, you might be better off spending the extra cash on the Z Ultra as it performed better in our tests. While vacuuming carpets isn't its thing, it still aced its mop tests.
Read our in-depth Narwal Freo Z Ultra review

How I tested the Narwal Freo Z10

• Used three times a week for six weeks
• Set up different custom routines and performed room cleans
• Tried different suction and waterflow settings

(Image credit: Sharmishta Sarkar / TechRadar)

As previously. mentioned, I tested the Narwal Freo Z10 in an apartment with mixed flooring, which gave me a good opportunity to assess both vacuuming and mopping prowess.

I tried all four suction modes and the different waterflow rates for both options, I set up various Shortcuts (or custom routines) and tried room cleaning too. I also delved into the settings in the app to adjust various options, particularly for the base station, and repeatedly checked the dirty-water tank for detergent use.

I also conducted TechRadar's standard tests for both vacuuming and mopping, scattering oats and tea from a bag onto both a carpet and tiles, and allowing a dollop of ketchup to dry.

I've previously tested several robot vacuums, including two in 2025 prior to the Freo Z10, giving me plenty of experience to compare performance and user experience.

Read more about how we test vacuums

[First tested July 2025]

President Trump has made some big shifts in U.S. policy on Russia's war with Ukraine lately.

In the course of two weeks, Trump halted and reinstated weapons to Ukraine and he began openly showing frustration with Russian president Vladimir Putin's continued military escalations.

Now, Trump has announced a deal with NATO to try to pressure Russia toward a ceasefire deal in just 50 days by threatening stiff tariffs and increased military aid to Ukraine.

President Trump seems to be taking a tougher stance against Russia, but will it make a difference, and will it last?

For sponsor-free episodes of Consider This, sign up for Consider This+ via Apple Podcasts or at plus.npr.org.

Email us at considerthis@npr.org.

• Sony is sharing more details on its NFL coaching headset
• The Sony NFL Coach's Headset will make its debut for the 2025 NFL season
• This one won't be purchasable by consumers, but promises best-in-class ANC like the WH-1000XM6

Sony is the official technology partner of the NFL – the National Football League – and we know that the tech giant has been working on a headset for coaches and other officials. We even got a first look at it back in January at the Consumer Electronics Show.

Now, though, Sony’s NFL Coach’s Headset is official, comes in three styles, and will be making its debut ahead of the 2025 season. And when the season does kick off later this year, you can expect 32 teams to be using the headset.

Unlike the latest Sony WH-1000XM6 for Sony, the NFL Coach’s Headset is tailor-made for, well, what the name describes. It’s designed from the ground up to work for coaches in a game environment, and that starts with connectivity. It doesn’t have Bluetooth onboard, but plugs into a special connectivity box that taps into the private network, powered by Verizon, for the NFL and the teams.

(Image credit: Sony)

There are also physical buttons or capacitive touch controls on the NFL Coach’s Headset. It also doesn’t have a rechargeable battery; rather, Sony is going old-school, possibly a quicker route than requiring a recharge by powering this headset with two AAA batteries. It’s also not carrying a specific IPX rating, but Sony did stress-test the unit for both extreme cold and extreme heat at NFL games.

This was done in environments where these weather conditions are recreated, as well as through live testing during NFL games this past season, including a frigid and snowy game at the Buffalo Bills' Highmark Stadium.

Sony’s past experience with various consumer headphones, including the WH-1000 series, will inform the design and other aspects, but it will likely focus on noise cancellation and pickup. It’s also not as straightforward as there are multiple design options for this headset, depending on the team and even individual preference.

Yes, it comes in a model with left and right earcups, but there are also two other options – just a left earcup and just a right earcup. All three, though, come with a microphone on a boom. Sony, however, is offering active noise cancellation on all three models.

Sony put this to the test in stadium environments where ambient crowd noise was measured at 100 decibels.

The resulting noise cancellation does, claims Sony, work effectively and has been tuned specifically for this use-case – here that means being able to hear communication while on the sidelines, but also for the microphone onboard using signal voice processing to pickup just the person speaking, and not the background chatter or even sound of the stadium.

Considering there are no buttons on the headphones, the microphone will automatically mute when the boom is raised. However, the belt pack, to which the headset will be plugged, will also have some manual controls.

Much of the focus here clearly went into making a durable headset that could withstand game after game use, though Sony confirmed with us that each team would have backups and that the overall feature set was purpose-crafted for each coaching staff.

Sony’s NFL Coach’s Headset will be worn during the 2025 season, which starts early September, but will make an early debut during the Hall of Fame game at the end of July. It, of course, does have some Sony branding front and center, after all, it’ll be shown on TV. I’m just curious if these will make it into a Beyond Sports simulcast – Homer rocking one could be neat.

As you might suspect with such a purpose-built product, there are no plans for a consumer release for the NFL Coach’s Headset. Those after a Sony headphone will need to consider the Sony WH-1000XM6, and you can read our full review here.

You might also like

• 'It's a lot of data, processed around 50 times per second' – Sony's Beyond Sports on how 'The Simpsons' Funday Football came to be
• As well as the Galaxy Ring 2, Samsung is also looking at smart earrings, necklaces, and other wearables
• 'I think you see the future first on Android' – Google's Android leader Sameer Samat

One performs better overall, but is the price difference enough to matter? Let's compare Samsung's flip phones for 2025.

The Grand Canyon Lodge is the only hotel on the park's North Rim, which is closed for the rest of the season due to wildfire risk. The hotel was already rebuilt once, after a kitchen fire in 1932.

Samsung's new book-style foldable has big upgrades on its predecessor -- so how do they stack up?

The Fed's $2.5 billion headquarters renovation is attracting mounting criticism from the Trump administration, which had been already attacking the central bank for not cutting interest rates.

(Image credit: Anna Moneymaker)

• Freepik launches unlimited AI video and image generation for Premium+ and Pro plans
• Model rollout will be gradual, but more are expected to be supported soon
• It could cost Freepik in the short term, but it'll result in long-term customers

Freepik is rolling out unlimited AI generation across the platform for paying users, meaning they won't be limited to caps or credits as is the case with many other rival platforms.

Premium+ and Pro account holders now get unlimited AI video and image generation as Freepik becomes one of the first major platforms to remove restrictions entirely.

Despite lifted limits, users will still be limited in other aspects - for example, video will be rendered in 768p resolution, using the MiniMax model from launch.

Freepik lifts AI limits

The company has promised to add further models as time goes on, with weekly launches planned for July 2025. Most will become unlimited, just like MiniMax, but some of the more powerful models like Veo 3 will be restricted.

"We decided to eliminate credits and offer unlimited generation because we understood that what holds users back is not the technology, but the frictions in the usage model," Freepik CEO Joaquin Cuenca explained.

Cuenca emphasized accessibility, creativity and mass adoption were the company's key drivers, rather than monetizing limitations, with the company expecting to absorb technological costs to promote long-term platform loyalty, ultimately leading to more sustainable revenue from long-term customers.

To get access to unlimited video and image generation, users will need to be on one of the two paying plans, starting at $24.50 per month.

"That's the real revolution of AI: not just in what it can do, but in how it is put at the service of people," the Cuenca added.

No details have been shared about upcoming models, but we do know that Premium+ and Pro plans get priority speed when using ChatGPT, Imagen 4 and Veo 3 compared with the cheaper and free plans. They also get early access to upcoming AI features, with top-tier Pro customers getting advanced AI models soonest.

You might also like

• Freepik launches new enterprise AI plan - but is it enough to tackle Adobe’s dominance in the field?
• We've listed the best Photoshop alternatives
• Check out the best AI tools

The Trump administration had appealed a decision that had directed it to stop gutting the U.S. Education Department and to reinstate many of the workers the government had laid off.

(Image credit: Win McNamee)

• Silverstone Seta H2 could be overkill for some, but it solves a very specific problem
• Storage density is the priority, and that comes with layout and thermal trade-offs
• Cable clutter and airflow chaos are inevitable when you chase maximum drive capacity

In a market full of flashy PC cases with glass panels, RGB lighting, and limited internal expandability, SilverStone’s newly unveiled Seta H2 case takes a far more practical approach focused on functionality.

Built as a full tower workstation case, the Seta H2 is about storage expansion rather than stylistic embellishments, and while its 540TB capacity might sound like overkill, this case makes it technically possible.

At a glance, it may look like a throwback, but beneath its plain surface lies the capacity to support what could be the largest HDD array in any consumer-grade tower case.

Not flashy, but engineered for scale

The case’s internal volume of 70 liters is used efficiently to accommodate up to 15 hard drives, and if each of these drives is 36TB, this enables a theoretical storage capacity of 540TB.

This configuration requires multiple removable brackets and cages, which allow users to mount a mix of 2.5-inch and 3.5-inch drives.

Additional 2.5-inch slots are hidden behind the motherboard tray and in various corners, suggesting this design caters to users who value storage density over airflow or clean cable layouts.

Enthusiasts considering this setup may find cooling to be a bottleneck, despite support for multiple fans and even large radiators.

Airflow becomes more complicated when 15 drives are tightly packed in the front, and those drives themselves are not exactly low-power or low-heat components.

Support for E-ATX and SSI-EEB motherboards makes the Seta H2 viable for enterprise or heavy workstation use.

The ability to fit long GPUs, up to 428mm, is impressive given the limited internal space, but installing a side radiator or using one of the drive brackets near the GPU can reduce clearance and make cooling and layout choices more difficult.

Whether the Seta H2 offers the best HDD setup is debatable, as power, heat, and cable management issues may limit its practical use.

With a starting price of around $216 or €200, this case is neither budget-friendly nor prohibitively expensive.

However, if you need the full 540TB capacity, a 36TB HDD like the Seagate Exos M 36TB is priced at $800.

At this rate, the total cost for 540TB could be over $12,000, depending on the models selected and current market conditions.

Via Techpowerup

You might also like

• These are the fastest SSDs you can buy right now
• Take a look at some of the best external hard drives on the market
• Good news: worldwide PC shipments are up. Bad news: it's due to the Windows 10 end of life and business upgrades