Now that Google is letting any developers bring their Android apps to the Chrome Web Store, the company needs to give developers a way to test their apps. It turns out the same tool lets anyone run (some) Android apps in Chrome… even if you’re using the Chrome web browser on Windows, OS X, or Linux rather than Chrome OS.
You have to jump through a few hoops, you can
only run one Android app at a time with this method, and not all apps will run properly. Google points out that ARC Welder is a beta/developer preview at this point (and that you can’t use it to publish apps to the Chrome Web Store).
Update: There’s a way to install and run multiple apps, but it’s a little complicated.
But if you don’t want to wait for developers of your favorite Android apps to bring them to Chrome, here’s how to try running the app yourself.
1. Install the ARC Welder Chrome app
ARC Welder is a free app available from the Chrome Web Store. Just click the “Add to Chrome” button and wait for the app to download and install.
Want to know more about ARC (Android Runtime for Chrome)? Google’s got a getting started page. Just keep in mind, it’s designed as a tool for developers, even though it’s available for anyone to use.
2. Get some Android APK files
The Google Play Store makes it easy to install apps on an Android device with a tap or two, but downloading the APK files to your computer can be a bit trickier. There are a few ways to do it though.
One of the easiest ways to grab APK files on a PC is to use a tool called Raccoon which lets you browse the Play Store from a PC and manually download apps.
If you have a rooted device you can also use a tool like Titanium Backup to save APK files for apps already installed on your phone or tablet to a backup folder. You can then manually copy them to your PC.
You can also find many APK files for popular Android apps at APK Mirror, a site run by the folks at Android Police.
Or you could try your luck searching the web or third-party app stores for APK files… just be careful when downloading APKs from unknown sources — you may be inadvertently downloading spyware or other malware along with your apps or games.
3. Use ARC Welder to load and run your Android apps
OK, so you’ve got a few APKs and you’ve got ARC Welder. Now what?
Open the Chrome app launcher, click the ARC Welder icon, and then hit the button that says “add your APK.”
The first time you do this you’ll need to choose a folder on your computer where your Android app will be stored. After that, just browse to your APK file and open it.
This’ll bring up a new window that says “Test your App.” Select your options (portrait or landscape mode, tablet, or phone form factor, maximized or window mode, etc) and then hit the “Launch App” button.
If everything works, a moment later your Android app will load and you can interact with it as if you were using an Android phone or tablet — although note that some touch-friendly apps might be harder than others to use with a keyboard and mouse.
Once an app is loaded this way, you can fire it up at any time from the Chrome launcher without first opening ARC Welder. But the next time you load an Android app it’ll overwrite the first one.
So how well does it work?
I tested a half dozen apps and found that the Songza, Hootsuite, and Business Calendar 2 apps all worked (somewhat) perfectly.
The Amazon Kindle app crashed every time I tried to load it, Hoopla didn’t load at all, and Google Play Books told me it wouldn’t run because Google Play Services wasn’t installed.
While Business Calendar loaded just fine and allowed me to create a new calendar, it unsurprisingly didn’t find any existing calendars since I was running the app on a non-Android device that doesn’t have any of my Google calendars associated with it.
If and when developers fully port their apps to run on Chrome, you probably won’t encounter these sorts of problems. But it’s worth noting that while you can try to run just about any Android app in Chrome now, not every app will work properly.
ToolHip Thank you for sharing ToolHippo
I used successfully ARC Welder to give my users the ability to use my WorkOnIT app on Windows under 8.1 – now when I upgraded my system to Windows 10 it no longer works – give continual crashes and it complains all the time about security certificates. Really a shame because under 8.1 it worked perfectly. Anyone else having big troubles with this with Windows 10?
I would like to see some way to makw these apps recogmise a touchscreen on like… a surface…
Does anybody tried it with his own app? I compiled my own app and after launching it the icon is displayed only
This is EXACTLY what we need. I want to see EVERY copy of Chrome include the ability to seamlessly run Android apps. This is the path towards truly making Android the universal runtime for end user applications.
Orrr developers can just create native Windows Apps, ohh my! have forgotten everyone hate Microsoft but wont stop using Windows.
I don’t get these developers you want android apps so bad on Windows and there is a way to make windows apps why not port them over? Instead of this one app at a time.
Exactly, Windows Universal Apps framework is so superior than this. This screams that Google is threatened by Microsoft new plan to conquer the app store game and cross platform applications. Because Windows 10 will not only run on PCs, it run on phones, tablets, xbox one’s, hololens, all types of wearable’s, and whole of the Internet of Things devices/senors e.g. from cars/vehicles to house appliances, smart homes/hotel automations, smart hospitals. This is not a tailored experience, its a app that fits one size fits all, which will not work perfectly, that will lead into bad app experiences across devices while Windows apps framework for 10 is scalable and tailored experience for each device, with no hard work of implementation and its an adaptive UI/UX platform. Wait, its already crushing this already, and BUILD 2015, will add even more insult to Google, when Visual Studio/Xaramin, Apache Cordova and other framework to get cross platform apps from other platforms like iOS/Android to Windows 10 easily in a tailored experience fitting modern design UI. This is nothing compared to what Microsoft is trying to achieve already.
I hope a lot of developers see App development from this point of view,
I mean it will be crazy as a developer for you not to be moved by that.
What MS is doing is incredible. I can’t wait to see how everything turns out.
What Qt has been doing for years is superior to everything out there including HTML5/Cordova as app platform. But much as the sheer weight of Microsoft will push Universal Apps as the be-all-end-all solution that it isn’t, sheer weight of Google might do the same for their own, if a bit inane, ideal.
Except that nobody wants to develop Windows Apps.
Yes there is a whole lot of sense in that.
I think you’ll find large numbers of developers are writing Windows software, as they have done for years.
The lack seems to be specifically the kind of apps written by non-software companies – website wrapper apps, or an app for a company’s services. The original point was that some mobile phones needed an application as a hacky workaround to cope with the small screen size. But now we have this odd situation where tablets get apps, but laptops are expected to use a website (or miss out).
ARCWelder gives application developers access not only to Windows desktop, but to all other platforms that can run ARCWelder. They don’t have to support yet another UI toolkit or yet another core sdk, which is what porting to Windows would involve. That is why this is more interesting than simply porting to Windows at this point: one set of updates to the application to play nicely with ARCWelder, and you get access to any platform ARCWelder runs on. No per-platform tweaking required.
Having a great OS and all the coolest development tools in the world won’t bring developers. (If that were the case, Solaris wouldn’t have drifted off into obscurity five years ago.) Market share will bring developers. If it’s a chicken-and-egg problem for Microsoft to enter these markets because their app base is too small, then they need to start incentivizing dev studios –*lots* of dev studios– to enter their application fold.
On the other hand, if Microsoft is willing to bleed money by keeping low-market-share products in the market for a long enough time, developers may eventually come to them anyway as the Android and iOS application market gets more crowded. Too much competition within a given application space will eventually push some developers to look for a less-crowded, if smaller, market.
Windows don’t have a low market share, the whole user base of MacOS is not up to Windows 8 & 8.1 alone.
The deciding factor right now is neither chicken nor egg its Windows 10, lets see if the whole world will upgrade to FREE Windows 10.
I could have been clearer about the context. I’m not talking about Microsoft’s share of the desktop market — which they absolutely dominate, no question– I’m talking about their share of the mobile/wearable/consumer-oriented embeddables market. Their market share in those areas is tiny compared to Android and iOS. Why is that the relevant market? Because that’s where the growth currently is, and it’s where the bulk of the (non-SAAS) developers are focusing their attention.
Almost everyone who develops desktops applications for a living already targets Windows. That’s not the issue. The issue is the horde of mobile developers who are steeped not in C++/C#/Qt/MFC/.Net, but in HTML5/AJAX/AJAJ/etc. ARCWelder gives those folks a relatively painless way to push their existing apps onto the desktop and the “in-between” space occupied by Chromebooks, without losing any time or ground in the mobile platform that is already their bread-and-butter. Shifting to mobile development for Windows carries a higher penalty in terms of initial time and capital.
Or they could use Qt and design properly responsive UIs in single toolkit, or develop web-based apps in something like Ionic and use knowledge of Angular to scale them beyond mobile and run them in node-webkit on desktop.
I think this is Google pushing Android SDK way, way beyond it’s proper use, but sheer amount of existing Android apps could make the idea viable, even if it’s tad disturbing.
They could, but that’s not the set of technologies the vast horde of mobile developers is using right now. I wish Qt had managed to invade the Android space much earlier.
yeah.. ever tried to use Qt on Android?
if your answer is no: don’t! isn’t worth the hassle.
Agree, though note this method is only intended for distributing to ChromeOS anyway (getting it to work on Windows only works for the test tool – although end users could use it, it’s not a very workable method for distribution).
Raccoon doesn’t seem to be able to find quite all of the applications on the Play Store. For instance, I searched for “iFruit” and it did not come up.
fdroid runs, but cant install software (for obvious reasons)
but still: you can download OpenSource-APKs from fdroid.org
How about AndrOpenOffice?
Lol, you could run chrome within chrome then 🙂
Felt like inception. Dream within dream. 🙂
A solid place to grab .apk files for the most popular apps is https://www.apkmirror.com/ , always up to date.
Good idea! That one totally slipped my mind while I was writing this post, but I’ve updated the article with a link.
Thanks a bunch for the link. Most APK repositories are extremely shady feeling. I like this one
Thank you – useful
Well if google actually does this right they could get rid of bluestacks/duos /console os/etc.
Bluestacks and other half-baked VM approaches yes, but hardly real OSs such ConsoleOS that runs on the bare metal. That said, it’s a dev tool so Joe Plinker should just stay away from it anyway.
How’s the performance compared to BlueStacks?
My assumption is that, as long as the app behaves properly in this runtime, it will probably perform much better than Bluestacks.
This method is essentially building the app to run in a runtime that is native to your CPU, rather than emulating the entire Android OS and ARM CPU. My experience with Bluestacks is limited, but that was my impression of how it worked.
While we’ll have to wait for developers to port their apps and have to wonder how functional they’ll be because a PC doesn’t have location based services, etc. that you’d find in pretty much every modern phone…
Mind that they’re just going to be running the apps, not the whole Android OS and that means some dependencies that some apps may rely on won’t be there… So, we’ll have to wait and see how well this goes…
Some (most?) Atom-based hybrid/2-in1’s have GPS and some also have cell tower communication. I’d guess much of the location dependent app functionality would be most important in such portable devices too.
You’re basically referring to tablets and phones, the GPS isn’t the only thing I was referring to anyway… Many apps basically require you to have the same components as a phone to work properly and there are other dependencies in the OS itself…
Like the article already shows, some apps are buggy, etc. and this is usually because it takes a bit more than just being able to run the app for it to work properly…
While most Chromebooks fall into the PC, rather than mobile category anyway, and this does also include those running Chrome on a regular PC as well as those are being included in these options…
Bluestacks is Android x86 (my children play android 3D games in decent quality on an old Core2Duo laptop), it’s not emulating ARM. The real question is where are we on the ART x86 front and according to the below, Intel is progressing nicely. Welder is Dalvik so it could be easily blown away if Bluestacks employs ART x86 performance wyse.
Comments are closed.