Vibe coding is programming by gut feel. You have an idea for a tool, a website, or a repetitive task you want to automate… but instead of enrolling in a coding boot camp or slogging through YouTube tutorials, you fire up an AI chatbot and have it write the code for you. This AI-driven approach to creative coding has exploded in popularity over the past year or two—chiefly because it works. People are building genuinely useful stuff without even knowing what a variable is. I’ve been vibe coding my own apps with AI and I’m absolutely loving it, despite being a relative newbie to coding. But most advice about AI-assisted programming still assumes you want to become a “real” programmer eventually. Because of that, the advice tends to steer you towards professional tools that are overkill for simple projects. You need a different—easier—approach. PROMOTIONLaunch in minutes by vibe coding Turn your ideas into reality with Hostinger Horizons. Build web apps, websites, and custom tools by chatting with the AI – no coding required. Get everything you need to launch and grow including web hosting, domain management and email. Start now Here’s what you really need to start vibe coding your own stuff right now, without any of the BS that only complicates matters. Use an AI chatbot, not an AI-assisted IDE Your first instinct might be to use an AI-assisted integrated development environment (IDE), one that you’ve heard is good for vibe coding, like Windsurf, Cursor, or even Claude Code (which technically isn’t an IDE, but that’s besides the point). All of these tools are built for AI coding, so they’re the best place to start, right? Well… no. These tools are designed to make highly productive programmers even more productive, and that means they assume a level of knowledge that amateur coders—like you and me—lack. Diving into these programs straight away is likely to leave you confused and discouraged. An AI-driven code editor can be powerful, but it’s overkill for an amateur vibe coder. This one starts at $15 per month after a two-week trial.Windsurf Not to mention that AI-powered IDEs are pretty darn expensive, too. While most offer a bit of free use, it’s easy for even uber-casual vibe coders to find themselves in need of a paid plan. Many such IDEs charge based on usage instead of a flat monthly rate, too. As an amateur vibe coder, you’re better off using an AI chatbot. The downside here is that you’ll need to copy-paste code back and forth between the chatbot and your code editor. It’s a bit of a nuisance, but it’s not too bad if your project is small. Choose your AI chatbot carefully When it comes to AI chatbots with programming proficiency, there are several you can choose from. All of the popular ones are good enough to get started, but even so, I have a couple recommendations. Don’t want to spend a dime? Try Google’s AI Studio. This chronically under-appreciated AI tool provides free and effectively unlimited access to Google’s latest Gemini Pro models, which many consider the best in the business without paying a cent. Google AI Studio does have a few downsides. Google will train on your data (most paid services promise they don’t), the interface isn’t the best, and there’s no desktop or mobile version. None of these are serious obstacles for amateur vibe coding, though. A screenshot of a quick data visualization tool I created using Claude.Matt Smith / Foundry But if you’re willing to pay a bit and want an attractive interface with desktop and mobile apps, I recommend Anthropic’s Claude Opus. I like Claude’s Artifacts feature, which breaks code out in a separate interface pane. It can even execute some code (like HTML or JavaScript) to provide a preview in Claude itself, and you can download code to your PC from this pane. Competitors have emulated it with similar features, but I think Anthropic’s implementation remains the best. Stick to popular programming languages Choosing a programming language is a major obstacle for both amateur and experienced programmers. There are hundreds to choose from, dozens of them are mature, and promoters of each tend to have strong opinions, so it’s easy to fall into analysis paralysis. When it comes to vibe coding, though, I recommend that you stick with languages that are both popular and forgiving. That means the usual suspects of JavaScript/TypeScript, Python, HTML, and perhaps C# if you’re feeling saucy. AI does best with these languages because AI models are trained on data from across the web. The more popular the language, the more examples there are to train on—and the more source data to work with, the better the generative results. As for how forgiving a particular language is, it’s an important aspect to consider because you’ll be quicker to get a program up and running even when it has bugs. Some languages, like Rust and C++, have strict syntax and architectural demands that can be a nightmare to debug. Veteran coders are likely to recoil at the idea of willingly accepting bugs in a program, but for beginners, it’s better to have a buggy-yet-functional program than a program that was abandoned because it never worked. Alternatively, you could let the AI choose your language for you. Describe the program you want to create and ask the chatbot to make it in whatever language seems most appropriate. Choose a simple code editor Once you have a chatbot and coding language picked, you’ll need to choose the software you use to edit code. You may feel pushed towards a full-blown programming IDE like Visual Studio, which is admittedly powerful but, again, designed for people who have some knowledge of programming. For casual vibe coding, I recommend a simpler code editor like Notepad++, Sublime Text, or UltraEdit. (My bias is towards Notepad++.) A screenshot of my vibe coded personal website in Notepad++.Matt Smith / Foundry Code editors lack the comprehensive terminal, debugging, and remote development features found in many IDEs. However, code editors still provide a fast, functional, and approachable interface that looks more like the barebones text editors you’ve likely used (e.g., Notepad), and they still provide handy features like code auto-completion. Think small (and reinvent the wheel) Scope creep is a common problem that all programmers face, and it can stop you cold. Big projects are more difficult than small projects, and vibe coding grows more difficult as the size of a project increases. So while it might be tempting to try vibe coding your way to a full-blown 3D game or world-class accounting software, you might want to hold your horses and start small. Very small.  A screenshot of me asking Claude about some enhancements to my tabletop game web app, which is relatively simple and small in scope.Matt Smith / Foundry Here are some great entry-level project examples for vibe coding: Python scripts that automate specific tasks, like scraping web pages for data or filtering documents by various criteria. A single-page personal website or app in HTML and JavaScript. A single-feature app programmed in Python or C#. My first vibe-coded project was a dice-rolling tool with text-to-speech capability for the tabletop strategy game Battletech. It can roll dice and announce the results of a roll, speeding up play. I also vibe-coded my own personal website, a single page of HTML with some JavaScript. Use AI to troubleshoot issues, answer questions, and generate documentation You might be surprised by how quickly AI can help you vibe code basic tools and programs. It’s so easy to pop out a quick Python script, you may find yourself doing it several times a day at one point. However, heavy reliance on vibe-coded projects can lead to another problem: losing track of how each script and program works, and becoming confused by dependencies. For example, Python programs typically don’t work right out of the box. You’ll need to install Python as well as the various libraries that are used to accomplish different things in a script. Not only do you have to figure that out, but if you ever move to another computer, you’ll also have to set up Python and those dependencies on that computer as well. Fortunately, AI is great at solving this problem. Just ask and it’ll provide you with instructions on how to install a new programming language and any related libraries on your computer. This will typically involve a handful of command-line prompts. I also recommend ending every vibe coding session by asking the AI to create documentation that teaches new users how to use the generated program. And when you use AI to modify existing scripts or programs, it’s a good idea to ask for a changelog of everything it does. Documentation will help you remember how your program works and help you get it working when you switch computers. Pick up Arduino and/or Raspberry Pi Python scripts and web pages can be useful, but they lack pizzazz… and that can be discouraging for any new vibe coder. If you want to try programming something that feels more tangible and exciting, then I recommend picking up an Arduino and/or Raspberry Pi. Going this route is likely to increase your starting costs and will make your vibe coding journey more complicated. You’ll need to buy some hardware and most likely a soldering iron. You’re also going to need to pick up some skills that AI can’t help with, like soldering. A Raspberry Pi Pico 2 W hooked up and ready to tinker with.Raspberry Pi While you can use AI to suggest Arduino and/or Raspberry Pi projects, I recommend picking a tutorial that already exists, like this Raspberry Pi Pico W weather station by ExplainingComputers. The tutorial explains the basics, of course, but the AI chatbot can further explain concepts that you don’t quite understand and handle edge cases. For example, I had a different temperature sensor than the one in the tutorial, so I had to wire it differently than how it was shown and make changes to the code to identify my sensor. With the help of AI, I was able to make the required changes in just 15 minutes. Further reading: Practical Raspberry Pi projects anyone can do PROMOTIONLaunch in minutes by vibe coding Turn your ideas into reality with Hostinger Horizons. Build web apps, websites, and custom tools by chatting with the AI – no coding required. Get everything you need to launch and grow including web hosting, domain management and email. Start now Read...

If you’re going to upgrade your computer, then you might as well get a mini PC these days because they’re a lot smaller, just as powerful, and more affordable. The Geekom AX8 Max is currently 23% off at Amazon, which means you can purchase it for $579 instead of the usual $749. The beautiful thing about this mini PC is that it’s not just going to be a good choice for work, but also for leisure since the built-in Radeon 780M graphics card will enable you to play some games. That’s one of the most powerful integrated graphics solutions around. While you can’t expect to play games with the graphics pushed at max level, you can still have some fun (as long as you temper your expectations!). You’ll also use the mini PC for work and your daily internet activities just fine. There’s a speedy AMD Ryzen 7 8745HS CPU on board and it comes loaded with a massive 32GB of DDR5 RAM. Now that’s a speedy combo that will serve you well day after day. It also doesn’t hurt that it features a 1TB SSD because that’s a ton of storage space! Plus, if you really want a speedier device, you can upgrade to 96GB of RAM and 4TB of storage without issue. Connectivity won’t be an issue with this mini PC because there are two USB4 type-C ports and two HDMI ports available, so you can hook up to four 8K screens. There are also tons of other ports for peripherals and data transfers, and even an audio jack if you want to take the old-fashioned approach to listening to music. So, snag this uber-powerful mini PC for $579 while this deal still lasts. Keep in mind this is a Prime deal, so you’ll need to have a Prime subscription to access it, which you can get with a 30-day free trial. Timing is perfect because Prime Day is coming July 8th to 11th with a gazillion discounts. Get a mini PC for 23% offBuy now at Amazon Read...

There are gentle migration strategies that can delay the move away from Windows 10, which Microsoft will stop supporting in October 2025. Technically (and maybe even psychologically), the easiest step is the hard one: get rid of Windows 10 and install an attractive, sustainable Linux desktop system in its place! 1. Preliminary considerations: Hardware and software Linux places no special demands on hardware, so there are generally no limitations when choosing a Linux system. Even more feature-rich desktop distributions like Ubuntu with the Gnome desktop, Kubuntu with KDE, or Linux Mint will run smoothly on any hardware that previously ran Windows 10. You only need to consider lighter distributions if the hardware was already noticeably struggling with Windows 10. Tip: For older notebooks with a mechanical 2.5-inch hard drive, the switch to Linux is a great opportunity to replace the old drive with an SSD. Delete and replace Windows 10: the easiest option technically for Linux installers (here Kubuntu with Calamares).IDG That way, you have a complete physical backup of your old Windows system on the mechanical hard drive, so you don’t need to worry about making a separate backup (see point two). To use this backup, you can use SATA-to-USB adapters, hard drive docking stations, or simply connect it to an internal SATA port in a spare PC (if available). Linux can replace Windows, but not everywhere. Switching is ideal and even improves security if you don’t need to run current or future software like Microsoft Office, Adobe Photoshop, Adobe InDesign, AutoCAD, or various Windows-only games. For all other tasks like internet browsing, chat, streaming, email, audio and video playback, photos, and Office file formats, Linux software works excellently. 2. Backup before installing Linux If you decide to replace Windows, make absolutely sure you save all important data. Missing anything means it can’t be recovered later. Backing up only the Windows home folder (“%userprofile%”) or a folder like “C:\Office” isn’t enough. Even copying key file types with commands like: xcopy /sc c:*.xls? e:\Backup or including “doc?”, “jpg”, and “txt” files is helpful but still doesn’t guarantee a complete backup. For Windows: The best method is to use the Disk2VHD tool. It creates VHD images that you can easily mount in Windows Explorer via Disk Management (diskmgmt.msc) by choosing “Action—Attach virtual hard drive.” For Linux: We recommend Rescuezilla, which creates IMG format images. You can mount these with tools like Gnome Disks (“Drives”) using “Attach Drive,” or sometimes even your file manager will work. Rescuezilla must be booted as a live external system, unlike Disk2VHD. 3. Deactivate license (retail versions) Before replacing Windows 10, make sure to deactivate your license. This usually frees up the product key, so you can use it on another PC later—or buy a cheaper PC without an OS. Most licenses, however, are OEM and tied to the current hardware. Only retail licenses can be activated on different machines. To deactivate, open a command prompt or PowerShell as an administrator and run: slmgr /upk This removes the activation. You can find your product key on your retail software packaging or in the confirmation email if you bought from the Microsoft Store. Tools like ShowKeyPlus (from the Microsoft Store) or this command can show your 25-digit key: wmic path SoftwareLicensingService get OA3xOriginalProductKey If no key appears, it’s probably not transferable. This also applies to keys on notebook stickers, which are usually OEM licenses. 4. Try out the Linux distribution in a live system Download the ISO file of the Linux distribution you want and copy it to a USB stick using a tool like Etcher or USB Imager. You can then boot your Windows computer from this USB by pressing F2, F8, F12, or Esc early on to access the BIOS boot menu (this varies by manufacturer, so just try the keys). In the boot menu, look for your USB stick listed under a name like “Intenso” or “Sandisk.” When the Linux live system starts, if offered, always choose “Try it out.” This lets you preview the desktop and test your hardware before installing.IDG Note: For the kind of installation that overwrites Windows, the boot mode (BIOS/MBR or UEFI/GPT) doesn’t matter. You can boot the USB with or without the “UEFI” label. Unlike Windows installers, most Linux ISO images (Ubuntu, Linux Mint, Zorin, Endeavour, etc) serve two purposes: they include an installer but also a full live system you can try out. Before choosing between “Try” and “Install,” we strongly recommend trying it first. This lets you explore the interface and, when you’re ready, starting the installer is just a click away. Even more important, this lets you check that key hardware like graphics, Ethernet, Wi-Fi, audio, printers, and Bluetooth work properly before making any changes. 5. Install Linux On the desktop of the live system, almost all Linux distributions show a shortcut to start the installation—for example, “Install Kubuntu 24.04 LTS.” The Ubuntu family uses two different installers: mostly Ubiquity, but some use Calamares. After selecting the language and keyboard, the installer will show a dialogue called “Installation Type” or “Partitions” (in Calamares). If it detects Windows on the primary hard drive, it will offer “Install Ubuntu alongside […]” or “Install in parallel” (Calamares) as the top, highlighted option. To install Linux as the only system, choose the “Erase hard drive” option. This will format the drive with the Linux Ext4 file system, completely delete Windows and its data, and install the selected Linux distribution. Clicking “Install now” or “Next” (Calamares) starts the installation, which then only asks for details to set up the first user account. Read...

If your PC won’t open Google Chrome, there may be a culprit: Microsoft. Google determined that Microsoft’s Family Safety app is blocking Chrome, though the company hasn’t discovered how or why. “For some users, Chrome is unable to run when Microsoft Family Safety is enabled,” Google support engineers posted on the company’s support blog. The issues have apparently been occurring for weeks. The glitch (?) was reported on by The Verge, who also reported that the problem can be resolved by temporarily renaming the Chrome executable to something else. (There’s nothing wrong with using the Google Chrome browser, of course.) Google also published the following workaround: Go to https://familysafety.microsoft.com or open the Family Safety mobile app. Select the child. To allow other browsers, Disable “Filter inappropriate websites” under the Edge tab, or Go to Windows tab ? Apps & Games ? unblock Chrome” Microsoft’s Family Safety allows a parent to manage screen time and apps used by their kids, and can either be accessed via the web or a dedicated app. Let’s hope that it’s just being a little aggressive, and Microsoft fixes it soon. Read...

Let’s say, as a thought experiment, that you’re a malware developer. You can choose to target specific groups of people to distribute your nefarious payloads. You might just go for a scattershot approach, but that’s less effective. You might aim for the technologically unsophisticated, or older people who aren’t as engaged. Or you can go for the lowest-hanging fruit: kids playing video games. That seems to be the choice for the latest batch of spyware, which is hiding in mods for the indefatigable Minecraft, now getting a predictable boost from a smash-hit kid’s movie. According to Check Point Research (via Bleeping Computer), a “multistage campaign” is going after Minecraft players looking for game mods on GitHub. The “Stargazers Ghost” network, allegedly a for-hire operation based in a Russian-speaking country, targets login information for the Minecraft game account, other third-party launchers, social media, and text app info. A second-stage program will try to steal more personal info in browser caches and other apps, with an emphasis on logins, passwords, and of course, cryptocurrency. The malware campaign has been spread across more than 500 GitHub repositories according to the report, and it’s a sneaky one, hiding in Minecraft Java installers to bypass many antivirus scans. Hiding this stuff on GitHub (which, like Minecraft itself, is owned by Microsoft) is particularly nefarious. While it’s mostly used for collaborative software development, downloads for software meant for end-users is now frequently hosted on GitHub itself. The technical interface and jargon might make it seem a little more trustworthy than a sketchy download site to an uncritical eye, or indeed, a child. GitHub does police its repositories for malware and other threats. But a centralized, company-based security team can often be overwhelmed by attackers based on numbers alone, to say nothing of a sophisticated campaign from a team of dedicated developers. To protect yourself or your kids when looking for Minecraft mods, Bleeping Computer recommends checking out GitHub page thoroughly, and testing mods on a “burner” account. Alternatively, I might suggest locking down a kid’s computer and telling them they have to stick to the official mods available in the “Bedrock Edition” of the game. Read...

DuckDuckGo has now extended the fraud protection in its privacy-centric browser to shield users from even more scams — without subjecting your browsing to additional tracking. Yay! In addition to blocking phishing sites, malware, and known online scams, the DuckDuck Go browser’s protection will now also work against fake online stores, fake crypto exchanges, and “scareware” sites that claim your PC is infected or blocked by the FBI, for example. The tool works by simply alerting the user if they click on a web page that could be potentially harmful. The website address is checked against a localized list of known threats from Netcraft that is updated every 20 minutes. Private people will appreciate how DDG handles the technical side of this. Unlike browsers that use Google Safe Browsing — such as Chrome, Safari, and Firefox — DuckDuckGo does all the checking without sharing data with a third party. The new feature is enabled by default without requiring an account. Further reading: I switched from Google to DuckDuckGo’s private search: 5 big takeaways Read...

In a country as vast as Australia, it’s critical to ensure that the benefits of access to reliable home internet aren’t just available to city dwellers. Read...

One of the best things about laptops nowadays is that they’re slim and thin, but this can also be their main downside because ports have been sacrificed on the altar of ultra-portability. Thankfully, a dock like this Ugreen Revodok Pro 313 can solve the problem just fine and it will only cost you $70 right now, instead of its usual $90. We actually really loved testing out this Ugreen dock, finding it has a great performance and a great price point. Our expert also appreciated its optimized build, although he didn’t really like how short the built-in cord is. Still, this is a device that got a 4.5-star rating from us and our Editors’ Choice award. If you’re wondering just how versatile the gadget is, you should know this is a 13-in-1 docking station that supports three displays. Yep, you read that right — you can connect up to three displays to your laptop, effectively turning it into a multitasker’s dream setup. It’s not just display ports and HDMIs either, because there are loads of USB-A ports too, an audio jack, as well as SD and TF card readers, which can obviously come in handy if you take tons of photos with a camera or if you want to access the videos your dashcam captured, for instance. Oh, and if you want to drop Wi-Fi and use an internet cable instead, there’s an Ethernet port available. So snag this one for $70 at Amazon while this deal is still available. Get a 22% discount on this 13-in-1 Ugreen docking stationBuy now at Amazon Read...

At a glanceExpert`s Rating Pros Intuitive web-based interface Robust security options Lots of business-ready capabilities Cons Business features won’t be needed by some users Requires annual commitment to get best value Our Verdict pdfFiller is an impressive PDF editor and document management solution with plenty to offer individuals and businesses. Price When Reviewed This value will show the geolocated pricing text for product undefined Best Pricing Today pdfFiller is a browser-based document management tool aimed at both individual and enterprise users. The end-to-end PDF solution allows you to do everything from editing to securing to sharing and storing your PDF files from its simple web interface. The app is designed like a desktop editor with a toolbar above the document pane and thumbnails down the left side that make it easy to manage multipage documents. Working with PDFs is impressively straightforward. Adding elements like text boxes, shapes, sticky notes, and even the date is as simple as clicking the appropriate icon and dragging that item to the page. Other tasks such as adding images and signatures require a few more steps, but typically employ a wizard to streamline the execution. Read on to learn more, then see our roundup of the best PDF editors for comparison. Editing and annotation To open a PDF, you can drag-and-drop the file directly onto pdfFiller’s home page. Alternately, you can upload files from your hard drive or from Google Drive, Dropbox, OneDrive, Box, a URL, or pdfFiller’s own online document library. You can also import a PDF from email by sending the file yourself or requesting it from a third party. You can simply drag elements like text boxes, shapes, and the date to your PDF. Michael Ansaldo/Foundry The editor enables you to work with PDFs much as you would with a Word document, allowing you to add, delete, and copy/paste text, change font style, size, and color; and so on. It also offers a standard set of markup tools. You can highlight, erase, or redact text; add sticky notes and comments; and scribble marginalia. A new “AI Replace” tool leverages ChatGPT to search across the PDF and automatically replace your target words or phrases while retaining original fonts, layout, and formatting. This makes batch updates — like changing company names, dates, or terms across a multi-page PDF — fast, accurate, and formatting-safe. Form hosting PDFs are often used to distribute contracts, questionnaires, and other types of forms. pdfFiller allows you to create, host, and edit these as easily as Word-style documents. You can also collect payments through PDF forms by linking a payment gateway. Document library One of pdfFiller’s most impressive features is its document library. If you can’t create the document you need from scratch, you can likely find it in this trove of government, legal, and business forms, any of which can be downloaded and customized to your needs. You can also save any document you create as a template for reuse. You can create PDFs from a library of government forms in pdfFiller’s document library. MIchael Ansaldo/Foundry One of pdfFiller’s most impressive features is its document library. Sharing Once your PDF is edited, you can securely share it with teammates through a link. You can also send it via email, text, fax, or USPS, or have it notarized directly from your pdfFiller account. Encryption and security pdfFiller offers robust encryption and security features for users that frequently handle sensitive data. Documents can be locked in an encrypted folder and require two-factor authentication for access. The editor offers dedicated HIPAA compliance settings to ensure patient healthcare information is protected according to standards outlined by the Health Insurance Portability and Accountability Act. It also helps keep track of your account activity with an Audit Trail feature that shows which forms you’ve opened and shared, what time you logged into and out of your pdfFiller account, and other user activity. How much is pdfFiller? The Basic plan is best suited to individuals. It provides standard PDF editing and annotation features, cloud storage for PDF files, and customer support within a day for $20 month-to-month or $8 a month with an annual commitment. The Plus plan adds the ability to create reusable templates; merge, rearrange, or add pages; and add basic fillable fields, plus support within an hour for $30 month-to-month or $12 a month with an annual commitment. The Premium plan includes all the features of the Basic and Plus plans and adds e-signature workflows, access to the U.S. Legal Forms Library, and other business-ready features, plus instant chat support. There’s a 30-day free trial period, which should plenty of time to determine which plan is right for you. Is pdfFiller worth it? pdfFiller is an incredibly versatile PDF editor with an uncommonly deep set of features. Not all of them will suit everyone, but fortunately, pdfFiller offers three subscription tiers that logically tailor features to different users. It’s a particularly good option for businesses, but anyone can benefit from pdfFiller’s comprehensive skill set. Editor’s note: Because online services are often iterative, gaining new features and performance improvements over time, this review is subject to change in order to accurately reflect the current state of the service. Any changes to text or our final review verdict will be noted at the top of this article. Read...

Proton VPN has quickly established itself as one of the best VPNs on the market thanks to its outstanding privacy practices, solid speeds, and top-tier feature set. The service offers a veritable treasure-trove of extra security features, especially if you opt for one of the higher-tier plans. Not all of these features are configured to work automatically though. But with a little bit of know-how you can have the VPN optimized in no time. Below I’ll show you the top five settings that I think everyone should change in Proton VPN to get the most from their subscription. Turn on the kill switch Proton VPN’s kill switch is a must-use feature that will keep your privacy intact in the event of an accidental disconnection.Sam Singleton This is something that I recommend for all VPN users. Turn on your kill switch and leave it engaged at all times. A kill switch will automatically disable your internet access should the VPN connection drop unexpectedly. This prevents your personal data from being exposed in the event that your VPN accidentally fails. It’s an easy, fail-safe way to ensure your privacy continues to remain intact the entire time you’re using a VPN. Proton VPN offers two versions of its killswitch feature: a Standard version that temporarily disconnects the internet when a VPN connection is lost, and an Advanced version that will only ever allow internet access when connected to the VPN. I personally use and recommend the Standard version as there are times when I want to surf the web or play games without the need for a VPN connection, but with the added security of the fail-safe auto-disconnect.  Either way, to turn on the kill switch feature in Proton VPN, just click on the “Kill switch” icon along the right side of the main screen. Toggle it on and select your preferred version. Activate Netshield Turn on Proton VPN’s NetShield to block ads, malware, and trackers while using the service. Sam Singleton Netshield is Proton VPN’s malware and ad- and tracker-blocker feature. It is offered with Proton’s Plus and Unlimited plans, but not the free version. While paying for separate antivirus or ad-blocker programs might provide more comprehensive coverage, if you’re just looking for the basics then Netshield has you covered for no additional cost. It’s also extremely easy to set up. To turn on Netshield simply click on the Netshield icon on the right side of the homescreen and toggle the feature on. Then choose whether you would like to block only malware or ads, malware, and trackers. I recommend blocking everything as ads and trackers are not only annoying, but may be secretly compromising your privacy or collecting and selling your data to third parties.  Set WireGuard to be your automatic protocol WireGuard is almost always the best protocol option for most people. Sam Singleton VPN protocols play a crucial role in the speed and security of your connection. They can be thought of as both the traffic laws and road map that guides your traffic through a VPN. Nearly every VPN provider offers up a handful of protocol options to choose from, but you’ll almost always want to stick with WireGuard. Not only does it provide the best combination of high speeds and strong security, but it’s also open source, meaning it’s constantly being vetted and checked for vulnerabilities. As such, I recommend that everyone choose WireGuard as their default protocol at all times. To change your protocol options in Proton VPN, open up the Settings menu from the homescreen, then click on Protocol. From here, simply select the WireGuard option as your default. You’ll likely notice that there are both WireGuard UDP and WireGuard TCP options available. UDP is a good choice when speeds matter, such as with streaming and gaming, but TCP is going to be better when reliability is paramount, such as with file transfers. Take some time to configure split tunneling Setting up split tunneling can pay huge dividends in optimizing speed and privacy for each app.Sam Singleton Split tunneling is one of the most underutilized features among VPN users and those who aren’t taking advantage of it are losing out. It allows you to pick and choose which apps or websites you wish to run through the encrypted VPN tunnel and which you wish to keep outside of it. Proton VPN comes with a really versatile split tunneling feature that is also easy to set up. From the homescreen, click on “Split tunneling” and then toggle it on to enable the feature. There are two versions of split tunneling here, Standard and Inverse. Standard will allow you to pick and choose which apps or IP addresses to exclude from the VPN connection—keeping them routed through your home IP instead. Inverse will automatically exclude everything and you will then choose which apps or IP addresses to run through the VPN connection instead. It’s worth taking some time here to optimize your VPN connection. Personally, I run my browser and work applications through the VPN, but keep my online videogames excluded. This way, I can protect my web surfing, get around streaming blocks, and keep my work data secure while simultaneously keeping my gaming latency as low as possible. Enable VPN Accelerator The VPN Accelerator feature with Proton is a handy way to increase connection speeds, especially over long distances. Sam Singleton Proton VPN comes with a unique and helpful feature called the VPN Accelerator. Through a series of design optimizations such as CPU load redistribution, a TCP flow control algorithm for reduced latency, and the use of bare-metal servers, it claims to increase connection speeds by up to 400 percent. While my tests unfortunately did not show speed increases that great, they did moderately improve with the feature turned on.  The feature is designed to work best at longer distances and when I connected to European servers from the U.S. I did notice significantly reduced latency compared to when the feature was turned off. Proton VPN has turned on VPN Accelerator by default, but just to be sure, you can check by going to the Settings menu from the homescreen and clicking on “VPN Accelerator.” Make sure it is toggled on and you should be good to go. There are practically no downsides to enabling VPN Accelerator—I’ve only read rumors that it may cause instability issues, but I’ve never personally experienced them nor seen any credible evidence to support these claims.  The feature works with all protocols and is available for all platforms. Related Content Proton VPN review NordVPN vs. Proton VPN: Which VPN is better in 2025? 9 things you realize when you first start using a VPN 5 VPN features you need to start using today Which VPN protocol should you use? It makes a big difference Read...