Google Chrome

The release announcement was originally scheduled for September 3, 2008, and a comic by Scott McCloud was to be sent to journalists and bloggers explaining the features within the new browser.[28] Copies intended for Europe were shipped early and German blogger Philipp Lenssen of Google Blogoscoped made a scanned copy of the 38-page comic available on his website after receiving it on September 1, 2008.[29][30] Google subsequently made the comic available on Google Books,[31] and mentioned it on their official blog along with an explanation for the early release.[32] The product was named "Chrome" as an initial development project code name, because it is associated with fast cars and speed. Google kept the development project name as the final release name, as a "cheeky" or ironic moniker, as one of the main aims was to minimize the user interface chrome.[33][34]

An early version of Chromium for Linux, explaining the difference between Chrome and Chromium

The browser was first publicly released, officially as a beta version,[35] on September 2, 2008, for Windows XP and newer, and with support for 43 languages, and later as a "stable" public release on December 11, 2008. On that same day, a CNET news item drew attention to a passage in the Terms of Service statement for the initial beta release, which seemed to grant to Google a license to all content transferred via the Chrome browser.[36] This passage was inherited from the general Google terms of service.[37] Google responded to this criticism immediately by stating that the language used was borrowed from other products, and removed this passage from the Terms of Service.[14]

Chrome quickly gained about 1% usage share.[32][38][39] After the initial surge, usage share dropped until it hit a low of 0.69% in October 2008. It then started rising again and by December 2008, Chrome again passed the 1% threshold.[40] In early January 2009, CNET reported that Google planned to release versions of Chrome for OS X and Linux in the first half of the year.[41] The first official Chrome OS X and Linux developer previews were announced on June 4, 2009,[42] with a blog post saying they were missing many features and were intended for early feedback rather than general use.[43] In December 2009, Google released beta versions of Chrome for OS X and Linux.[44][45] Google Chrome 5.0, announced on May 25, 2010, was the first stable release to support all three platforms.[46]

Chrome was one of the twelve browsers offered on BrowserChoice.eu to European Economic Area users of Microsoft Windows in 2010.[47]

Chrome was assembled from 25 different code libraries from Google and third parties such as Mozilla's Netscape Portable Runtime, Network Security Services, NPAPI (dropped as of version 45),[48] Skia Graphics Engine, SQLite, and a number of other open-source projects.[49] The V8 JavaScript virtual machine was considered a sufficiently important project to be split off (as was Adobe/Mozilla's Tamarin) and handled by a separate team in Denmark coordinated by Lars Bak. According to Google, existing implementations were designed "for small programs, where the performance and interactivity of the system weren't that important", but web applications such as Gmail "are using the web browser to the fullest when it comes to DOM manipulations and JavaScript", and therefore would significantly benefit from a JavaScript engine that could work faster.

Chrome initially used the WebKit rendering engine to display web pages. In 2013, they forked the WebCore component to create their own layout engine Blink. Based on WebKit, Blink only uses WebKit's "WebCore" components, while substituting other components, such as its own multi-process architecture, in place of WebKit's native implementation.[17] Chrome is internally tested with unit testing, automated testing of scripted user actions, fuzz testing, as well as WebKit's layout tests (99% of which Chrome is claimed to have passed), and against commonly accessed websites inside the Google index within 20–30 minutes.[31] Google created Gears for Chrome, which added features for web developers typically relating to the building of web applications, including offline support.[31] Google phased out Gears as the same functionality became available in the HTML5 standards.[50]

In March 2011, Google introduced a new simplified logo to replace the previous 3D logo that had been used since the project's inception. Google designer Steve Rura explained the company reasoning for the change: "Since Chrome is all about making your web experience as easy and clutter-free as possible, we refreshed the Chrome icon to better represent these sentiments. A simpler icon embodies the Chrome spirit – to make the web quicker, lighter, and easier for all."[51]

On January 11, 2011, the Chrome product manager, Mike Jazayeri, announced that Chrome would remove H.264 video codec support for its HTML5 player, citing the desire to bring Google Chrome more in line with the currently available open codecs available in the Chromium project, which Chrome is based on.[52] Despite this, on November 6, 2012, Google released a version of Chrome on Windows which added hardware-accelerated H.264 video decoding.[53] In October 2013, Cisco announced that it was open-sourcing its H.264 codecs and would cover all fees required.[54]

On February 7, 2012, Google launched Google Chrome Beta for Android 4.0 devices.[55] On many new devices with Android 4.1 and later preinstalled, Chrome is the default browser.[56] In May 2017, Google announced a version of Chrome for augmented reality and virtual reality devices.[57]

The results of the Acid3 test on Google Chrome 4.0

The first release of Google Chrome passed both the Acid1 and Acid2 tests. Beginning with version 4.0, Chrome has passed all aspects of the Acid3 test.[62]

As of May 2011, Chrome has very good support for JavaScript/ECMAScript according to Ecma International's ECMAScript standards conformance Test 262[63] (version ES5.1 May 18, 2012). This test reports as the final score the number of tests a browser failed; hence lower scores are better. In this test, Chrome version 37 scored 10 failed/11,578 passed. For comparison, Firefox 19 scored 193 failed/11,752 passed and Internet Explorer 9 has a score of 600+ failed, while Internet Explorer 10 has a score of 7 failed.

In 2011, on the official CSS 2.1 test suite by standardization organization W3C, WebKit, the Chrome rendering engine, passes 89.75% (89.38% out of 99.59% covered) CSS 2.1 tests.[64]

On the HTML5 web standards test, Chrome 41 scores 518 out of 555 points, placing it ahead of the five most popular desktop browsers.[65][66] Chrome 41 on Android scores 510 out of 555 points.[67][68][69] Chrome 44 scores 526, only 29 points less than the maximum score.[70]

Google Chrome logos

2D motif from March 2011 until October 2015

Material Design motif used from September 2014 onward for mobile versions and October 2015 onward for desktop versions

New Google Chrome logo from 2022. It has increased brightness and clarity compared to the previous logo.

By default, the main user interface includes back, forward, refresh/cancel and menu buttons. A home button is not shown by default, but can be added through the Settings page to take the user to the new tab page or a custom home page.[71]

Tabs are the main component of Chrome's user interface and have been moved to the top of the window rather than below the controls. This subtle change contrasts with many existing tabbed browsers which are based on windows and contain tabs. Tabs, with their state, can be transferred seamlessly between window containers by dragging. Each tab has its own set of controls, including the Omnibox.[31]

The Omnibox is a URL box that combines the functions of both the address bar and search box. If a user enters the URL of a site previously searched from, Chrome allows pressing Tab to search the site again directly from the Omnibox. When a user starts typing in the Omnibox, Chrome provides suggestions for previously visited sites (based on the URL or in-page text), popular websites (not necessarily visited before – powered by Google Instant), and popular searches. Although Instant can be turned off, suggestions based on previously visited sites cannot be turned off. Chrome will also autocomplete the URLs of sites visited often.[31] If a user types keywords into the Omnibox that do not match any previously visited websites and presses enter, Chrome will conduct the search using the default search engine.

One of Chrome's differentiating features is the New Tab Page, which can replace the browser home page and is displayed when a new tab is created. Originally, this showed thumbnails of the nine most visited websites, along with frequent searches, recent bookmarks, and recently closed tabs; similar to Internet Explorer and Firefox with Google Toolbar, or Opera's Speed Dial.[31] In Google Chrome 2.0, the New Tab Page was updated to allow users to hide thumbnails they did not want to appear.[72]

Starting in version 3.0, the New Tab Page was revamped to display thumbnails of the eight most visited websites. The thumbnails could be rearranged, pinned, and removed. Alternatively, a list of text links could be displayed instead of thumbnails. It also features a "Recently closed" bar that shows recently closed tabs and a "tips" section that displays hints and tricks for using the browser.[73] Starting with Google Chrome 3.0, users can install themes to alter the appearance of the browser.[74] Many free third-party themes are provided in an online gallery,[75] accessible through a "Get themes" button in Chrome's options.[76]

Chrome includes a bookmarks submenu that lists the user's bookmarks, provides easy access to Chrome's Bookmark Manager, and allows the user to toggle a bookmarks bar on or off.

On January 2, 2019, Google introduced Native Dark Theme for Chrome on Windows 10.[77]

In 2023, it was announced that Chrome would be completely revamped, using Google's Material You design language, the revamp would include more rounded corners, Chrome colors being swapped out for a similar dynamic color system introduced in Android 12, a revamped address bar, new icons and tabs, and a more simplified 3 dot menu.[78]

Starting with Google Chrome 4.1 the application added a built-in translation bar using Google Translate. Language translation is currently available for 52 languages.[79] When Chrome detects a foreign language other than the user's preferred language set during the installation time, it asks the user whether or not to translate.

Chrome allows users to synchronize their bookmarks, history, and settings across all devices with the browser installed by sending and receiving data through a chosen Google Account, which in turn updates all signed-in instances of Chrome. This can be authenticated either through Google credentials, or a sync passphrase.

For web developers, Chrome features an element inspector which allows users to look into the DOM and see what makes up the webpage.[80]

Chrome has special URLs that load application-specific pages instead of websites or files on disk. Chrome also has a built-in ability to enable experimental features. Originally called about:labs, the address was changed to about:flags to make it less obvious to casual users.[81][82]

The desktop edition of Chrome is able to save pages as HTML with assets in a "_files" subfolder, or as unprocessed HTML-only document. It also offers an option to save in the MHTML format.[83]

Chrome allows users to make local desktop shortcuts that open web applications in the browser. The browser, when opened in this way, contains none of the regular interface except for the title bar, so as not to "interrupt anything the user is trying to do". This allows web applications to run alongside local software (similar to Mozilla Prism and Fluid).[31]

This feature, according to Google, would be enhanced with the Chrome Web Store, a one-stop web-based web applications directory which opened in December 2010.[84][85]

In September 2013, Google started making Chrome apps "For your desktop". This meant offline access, desktop shortcuts, and less dependence on Chrome—apps launch in a window separate from Chrome, and look more like native applications.[86]

Announced on December 7, 2010, the Chrome Web Store allows users to install web applications as extensions to the browser, although most of these extensions function simply as links to popular web pages or games, some of the apps like Springpad do provide extra features like offline access. The themes and extensions have also been tightly integrated into the new store, allowing users to search the entire catalog of Chrome extras.[87]

The Chrome Web Store was opened on February 11, 2011, with the release of Google Chrome 9.0.[88]

Browser extensions are able to modify Google Chrome. They are supported by the browser's desktop edition.[89] These extensions are written using web technologies like HTML, JavaScript, and CSS.[90] They are distributed through Chrome Web Store,[91] initially known as the Google Chrome Extensions Gallery.[89]

On September 9, 2009, Google enabled extensions by default on Chrome's developer channel, and provided several sample extensions for testing.[92] In December, the Google Chrome Extensions Gallery beta began with approximately 300 extensions.[45][93] It was launched on January 25, 2010, along with Google Chrome 4.0, containing approximately 1500 extensions.[94]

In 2014, Google started preventing some Windows users from installing extensions not hosted on the Chrome Web Store.[95][96] The following year Google reported a "75% drop in customer support help requests for uninstalling unwanted extensions" which led them to expand this restriction to all Windows and Mac users.[97]

The JavaScript virtual machine used by Chrome, the V8 JavaScript engine, has features such as dynamic code generation, hidden class transitions, and precise garbage collection.[31]

In 2008, several websites performed benchmark tests using the SunSpider JavaScript Benchmark tool as well as Google's own set of computationally intense benchmarks, which include ray tracing and constraint solving.[101] They unanimously reported that Chrome performed much faster than all competitors against which it had been tested, including Safari (for Windows), Firefox 3.0, Internet Explorer 7, Opera, and Internet Explorer 8.[102][103][104][60]{[105][106] However, on October 11, 2010, independent tests of JavaScript performance, Chrome has been scoring just behind Opera's Presto engine since it was updated in version 10.5.[107]

On September 3, 2008, Mozilla responded by stating that their own TraceMonkey JavaScript engine (then in beta), was faster than Chrome's V8 engine in some tests.[108][109][110] John Resig, Mozilla's JavaScript evangelist, further commented on the performance of different browsers on Google's own suite, commenting on Chrome's "decimating" of the other browsers, but he questioned whether Google's suite was representative of real programs. He stated that Firefox 3.0 performed poorly on recursion-intensive benchmarks, such as those of Google, because the Mozilla team had not implemented recursion-tracing yet.[111]

Two weeks after Chrome's launch in 2008, the WebKit team announced a new JavaScript engine, SquirrelFish Extreme,[112] citing a 36% speed improvement over Chrome's V8 engine.[113][114][115]

Like most major web browsers, Chrome uses DNS prefetching to speed up website lookups,[80] as do other browsers like Firefox,[116] Safari,[117] Internet Explorer (called DNS Pre-resolution),[118] and in Opera as a UserScript (not built-in).[119]

Chrome formerly used their now-deprecated SPDY protocol instead of only HTTP[120][121] when communicating with servers that support it, such as Google services, Facebook, Twitter. SPDY support was removed in Chrome version 51. This was due to SPDY being replaced by HTTP/2, a standard that was based upon it.

In November 2019, Google said it was working on several "speed badging" systems that let visitors know why a page is taking time to show up. The variations include simple text warnings and more subtle signs that indicate a site is slow. No date has been given for when the badging system will be included with the Chrome browser.[122]

Chrome formerly supported a Data Saver feature for making pages load faster called Lite Mode.[123] Previously, Chrome engineers Addy Osmani and Scott Little announced Lite Mode would automatically lazy-load images and iframes for faster page loads.[124] Lite Mode was switched off in Chrome 100, citing a decrease in mobile data costs for many countries.[125]

Chrome periodically retrieves updates of two blacklists (one for phishing and one for malware), and warns users when they attempt to visit a site flagged as potentially harmful. This service is also made available for use by others via a free public API called "Google Safe Browsing API".[31]

Chrome uses a process-allocation model to sandbox tabs.[126] Using the principle of least privilege, each tab process cannot interact with critical memory functions (e.g. OS memory, user files) or other tab processes – similar to Microsoft's "Protected Mode" used by Internet Explorer 9 or greater. The Sandbox Team is said to have "taken this existing process boundary and made it into a jail". This enforces a computer security model whereby there are two levels of multilevel security (user and sandbox) and the sandbox can only respond to communication requests initiated by the user.[127] On Linux sandboxing uses the seccomp mode.[128][129]

In January 2015, TorrentFreak reported that using Chrome when connected to the internet using a VPN can be a serious security issue due to the browser's support for WebRTC.[130]

On September 9, 2016, it was reported that starting with Chrome 56, users will be warned when they visit insecure HTTP websites to encourage more sites to make the transition to HTTPS.[131]

On December 4, 2018, Google announced its Chrome 71 release with new security features, including a built-in ad featuring system. In addition, Google also announced its plan to crack down on websites that make people involuntarily subscribe to mobile subscription plans.[132]

On September 2, 2020, with the release of Chrome 85, Google extended support for Secure DNS in Chrome for Android. DNS-over-HTTPS (DoH), was designed to improve safety and privacy while browsing the web. Under the update, Chrome automatically switches to DNS-over-HTTPS (DoH), if the current DNS provider supports the feature.[133]

Since 2008, Chrome has been faulted for not including a master password to prevent casual access to a user's passwords. Chrome developers have indicated that a master password does not provide real security against determined hackers and have refused to implement one. Bugs filed on this issue have been marked "WontFix".[134][135] As of February 2014, Google Chrome asks the user to enter the Windows account password before showing saved passwords.[136]

On Linux, Google Chrome/Chromium can store passwords in three ways: GNOME Keyring, KWallet or plain text. Google Chrome/Chromium chooses which store to use automatically, based on the desktop environment in use.[137] Passwords stored in GNOME Keyring or KWallet are encrypted on disk, and access to them is controlled by dedicated daemon software. Passwords stored in plain text are not encrypted. Because of this, when either GNOME Keyring or KWallet is in use, any unencrypted passwords that have been stored previously are automatically moved into the encrypted store. Support for using GNOME Keyring and KWallet was added in version 6, but using these (when available) was not made the default mode until version 12.

As of version 45, the Google Chrome password manager is no longer integrated with Keychain, since the interoperability goal is no longer possible.[138]

No security vulnerabilities in Chrome were exploited in the three years of Pwn2Own from 2009 to 2011.[139] At Pwn2Own 2012, Chrome was defeated by a French team who used zero day exploits in the version of Flash shipped with Chrome to take complete control of a fully patched 64-bit Windows 7 PC using a booby-trapped website that overcame Chrome's sandboxing.[140]

Chrome was compromised twice at the 2012 CanSecWest Pwnium.[140][141][142] Google's official response to the exploits was delivered by Jason Kersey, who congratulated the researchers, noting "We also believe that both submissions are works of art and deserve wider sharing and recognition."[143] Fixes for these vulnerabilities were deployed within 10 hours of the submission.[144][145]

A significant number of security vulnerabilities in Chrome occurred in the Adobe Flash Player. For example, the 2016 Pwn2Own successful attack on Chrome relied on four security vulnerabilities. Two of the vulnerabilities were in Flash, one was in Chrome, and one was in the Windows kernel.[146] In 2016, Google announced that it was planning to phase out Flash Player in Chrome, starting in version 53. The first phase of the plan was to disable Flash for ads and "background analytics", with the ultimate goal of disabling it completely by the end of the year, except on specific sites that Google has deemed to be broken without it. Flash would then be re-enabled with the exclusion of ads and background analytics on a site-by-site basis.[147]

Leaked documents published by WikiLeaks, codenamed Vault 7 and dated from 2013 to 2016, detail the capabilities of the CIA, such as the ability to compromise web browsers (including Google Chrome).[148][149]

Google introduced download scanning protection in Chrome 17.[150] In February 2018, Google introduced an ad blocking feature based on recommendations from the Interactive Advertising Bureau. Sites that employ invasive ads are given a 30-day warning, after which their ads will be blocked.[151] Consumer Reports recommended users install dedicated ad-blocking tools instead, which offer increased security against malware and tracking.[152]

• Chrome supported, up to version 45, plug-ins with the Netscape Plugin Application Programming Interface (NPAPI),[153] so that plug-ins (for example Adobe Flash Player) run as unrestricted separate processes outside the browser and cannot be sandboxed as tabs are. ActiveX is not supported.[153] Since 2010, Adobe Flash has been integral to Chrome and does not need be installed separately. Flash is kept up to date as part of Chrome's own updates.[154] Java applet support was available in Chrome with Java 6 update 12 and above.[155] Support for Java under OS X was provided by a Java Update released on May 18, 2010.[156]
• On August 12, 2009, Google introduced a replacement for NPAPI that is more portable and more secure[157] called Pepper Plugin API (PPAPI).[158] The default bundled PPAPI Flash Player (or Pepper-based Flash Player) was available on ChromeOS first, then replaced the NPAPI Flash Player on Linux from Chrome version 20, on Windows from version 21 (which also reduced Flash crashes by 20%),[159] and eventually came to OS X at version 23.[160]
• On September 23, 2013, Google announced that it would be deprecating and then removing NPAPI support. NPAPI support was removed from Linux in Chrome release 35.[161] NPAPI plugins like Java can no longer work in Chrome (but there are workarounds for Flash by using PPAPI Flash Player on Linux including for Chromium).[162]
• On April 14, 2015, Google released Chrome v42, disabling the NPAPI by default. This makes plugins that do not have a PPAPI plugin counterpart incompatible with Chrome, such as Java, Silverlight and Unity. However, NPAPI support could be enabled through the chrome://flags menu, until the release of version 45 on September 1, 2015, that removed NPAPI support entirely.[163]

Google Chrome Incognito mode message

The private browsing feature called Incognito mode prevents the browser from locally storing any history information, cookies, site data, or form inputs.[164] Downloaded files and bookmarks will be stored. In addition, user activity is not hidden from visited websites or the Internet service provider.[165]

Incognito mode is similar to the private browsing feature in other web browsers. It does not prevent saving in all windows: "You can switch between an incognito window and any regular windows you have open. You'll only be in incognito mode when you're using the incognito window".[166]

The iOS version of Chrome also supports the optional ability to lock incognito tabs with Face ID, Touch ID or the device's passcode.[167]

In June 2015, the Debian developer community discovered that Chromium 43 and Chrome 43 were programmed to download the Hotword Shared Module, which could enable the OK Google voice recognition extension, although by default it was "off". This raised privacy concerns in the media.[168][169] The module was removed in Chrome 45, which was released on September 1, 2015, and was only present in Chrome 43 and 44.[170][171]

Chrome sends details about its users and their activities to Google through both optional and non-optional user tracking mechanisms.[172][173]

Some of the tracking mechanisms can be optionally enabled and disabled through the installation interface[174] and through the browser's options dialog.[175] Unofficial builds, such as SRWare Iron, seek to remove these features from the browser altogether.[176] The RLZ feature is not included in the Chromium browser either.[177]

In March 2010, Google devised a new method to collect installation statistics: the unique ID token included with Chrome is now used for only the first connection that Google Update makes to its server.[178]

The optional suggestion service included in Google Chrome has been criticized because it provides the information typed into the Omnibox to the search provider before the user even hits return. This allows the search engine to provide URL suggestions, but also provides them with web use information tied to an IP address.[179]

Chrome previously was able to suggest similar pages when a page could not be found. For this, in some cases Google servers were contacted.[180] The feature has since been removed.

A 2019 review by Washington Post technology columnist Geoffrey A. Fowler found that in a typical week of browsing, Chrome allowed thousands more cookies to be stored than Mozilla Firefox. Fowler pointed out that because of its advertising businesses, despite the privacy controls it offers users, Google is a major producer of third-party cookies and has a financial interest in collecting user data; he recommended switching to Firefox, Apple Safari, or Chromium-based Brave.[181]

In January 2021, Google stated it was making progress on developing privacy-friendly alternatives which would replace third-party cookies currently being used by advertisers and companies to track browsing habits. Google then promised to phase out the use of cookies in their web-browser in 2022, implementing their FLoC technology instead. The announcement triggered antitrust concerns from multiple countries for abusing the Chrome browser's market monopoly, with the U.K.'s Competition and Markets Authority and the European Commission both opening formal probes.[186][187][188][189] The FLoC proposal also drew criticism from DuckDuckGo, Brave, and the Electronic Frontier Foundation for misrepresenting its ability to track users online.[190][191][192][193]

On January 25, 2022, Google announced it had killed off development of its FLoC technologies and proposed the new Topics API to replace it. Topics is similarly intended to replace cookies, using one's weekly web activity to determine a set of five interests. Topics is supposed to refresh every three weeks, changing the type of ads served to the user and not retaining the gathered data.[194][195]

In February 2012, Google announced that Chrome would implement the Do Not Track (DNT) standard to inform websites the user's desire not to be tracked. The protocol was implemented in version 23. In line with the W3's draft standard for DNT,[196] it is turned off by default in Chrome.[197]

Screenshot of a Chrome browser crash

A multi-process architecture is implemented in Chrome where, by default, a separate process is allocated to each site instance and plugin.[198] This procedure is termed process isolation,[199] and raises security and stability by preventing tasks from interfering with each other. An attacker successfully gaining access to one application gains access to no others,[200] and failure in one instance results in a Sad Tab screen of death, similar to the well-known Sad Mac, but only one tab crashes instead of the whole application. This strategy exacts a fixed per-process cost up front, but results in less memory bloat over time as fragmentation is confined to each instance and no longer needs further memory allocations.[31] This architecture was later adopted in Safari[201] and Firefox.[202]

Chrome includes a process management utility called Task Manager which lets users see what sites and plugins are using the most memory, downloading the most bytes and overusing the CPU and provides the ability to terminate them.[203] Chrome Version 23 ensures its users an improved battery life for the systems supporting Chrome's GPU accelerated video decoding.[204][53]

The first production release on December 11, 2008 marked the end of the initial Beta test period and the beginning of production. Shortly thereafter, on January 8, 2009, Google announced an updated release system with three channels: Stable (corresponding to the traditional production), Beta, and Developer preview (also called the "Dev" channel). Where there were before only two channels: Beta and Developer, now there were three. Concurrently, all Developer channel users were moved to the Beta channel along with the promoted Developer release. Google explained that now the Developer channel builds would be less stable and polished than those from the initial Google Chrome's Beta period. Beta users could opt back to the Developer channel as desired.

Each channel has its own release cycle and stability level. The Stable channel updated roughly quarterly, with features and fixes that passed "thorough" testing in the Beta channel. Beta updated roughly monthly, with "stable and complete" features migrated from the Developer channel. The Developer channel updated once or twice per week and was where ideas and features were first publicly exposed "(and sometimes fail) and can be very unstable at times". [Quoted remarks from Google's policy announcements.][205][206][207]

Google Chrome Canary application icon

On July 22, 2010, Google announced it would ramp up the speed at which it releases new stable versions; the release cycles were shortened from quarterly to six weeks for major Stable updates.[208] Beta channel releases now come roughly at the same rate as Stable releases, though approximately one month in advance, while Dev channel releases appear roughly once or twice weekly, allowing time for basic release-critical testing.[209] This faster release cycle also brought a fourth channel: the "Canary" channel, updated daily from a build produced at 09:00 UTC from the most stable of the last 40 revisions.[210] The name refers to the practice of using canaries in coal mines, so if a change "kills" Chrome Canary, it will be blocked from migrating down to the Developer channel, at least until fixed in a subsequent Canary build.[211] Canary is "the most bleeding-edge official version of Chrome and somewhat of a mix between Chrome dev and the Chromium snapshot builds". Canary releases run side by side with any other channel; it is not linked to the other Google Chrome installation and can therefore run different synchronization profiles, themes, and browser preferences. This ensures that fallback functionality remains even when some Canary updates may contain release-breaking bugs.[212] It does not natively include the option to be the default browser, although on Windows and OS X it can be set through System Preferences. Canary was Windows-only at first; an OS X version was released on May 3, 2011.[213]

The Chrome beta channel for Android was launched on January 10, 2013; like Canary, it runs side by side with the stable channel for Android.[214][215] Chrome Dev for Android was launched on April 29, 2015.[216]

All Chrome channels are automatically distributed according to their respective release cycles. The mechanism differs by platform. On Windows, it uses Google Update, and auto-update can be controlled via Group Policy.[217] Alternatively, users may download a standalone installer of a version of Chrome that does not auto-update.[218][219] On OS X, it uses Google Update Service, and auto-update can be controlled via the OS X "defaults" system.[220] On Linux, it lets the system's normal package management system supply the updates. This auto-updating behavior is a key difference from Chromium, the non-branded open-source browser which forms the core of Google Chrome. Because Chromium also serves as the pre-release development trunk for Chrome, its revisions are provided as source code and buildable snapshots are produced continuously with each new commit, requiring users to manage their own browser updates.[221]

In March 2021, Google announced that starting with Chrome 94 in the third quarter of 2021, Google Chrome Stable releases will be made every four weeks, instead of six weeks as they have been since 2010. Also, Google announced a new release channel for system administrators and browser embedders with releases every eight weeks.[222]

Releases are identified by a four-part version number, e.g. 42.0.2311.90 (Windows Stable release April 14, 2015[223]). The components are major.minor.build.patch.[224][225]

Major.minor reflects scheduling policy

Build.patch identifies content progression

• Major represents a product release. These are scheduled 7–8 per year, unlike other software systems where the major version number updates only with substantial new content.
• Minor is usually 0. References to version 'x' or 'x.0', e.g. 42.0, refer to this major.minor designation.
• Build is ever increasing. For a release cycle, e.g. 42.0, there are several builds in the Canary and Developer period. The last build number from Developer is kept throughout Beta and Stable and is locked with the major.minor for that release.
• Patch resets with each build, incrementing with each patch. The first patch is 0, but usually the first publicly released patch is somewhat higher. In Beta and Stable, only patch increments.

Chromium and Chrome release schedules are linked through Chromium (Major) version Branch Point dates, published annually.[224] The Branch Points precede the final Chrome Developer build (initial) release by 4 days (nearly always) and the Chrome Stable initial release by roughly 53 days.[226]

Example: The version 42 Branch Point was February 20, 2015.[224] Developer builds stopped advancing at build 2311 with release 42.0.2311.4 on February 24,[227] 4 days later. The first Stable release, 42.0.2311.90, was April 14, 2015,[223] 53 days after the Branch Point.

Chrome supports color management by using the system-provided ICC v2 and v4 support on macOS, and from version 22 supports ICC v2 profiles by default on other platforms.[228]

In Chrome, when not connected to the Internet and an error message displaying "No internet" is shown, on the top, an "8-bit" Tyrannosaurus rex is shown, but when pressing the space bar on a keyboard, mouse-clicking on it or tapping it on touch devices, the T-Rex instantly jumps once and dashes across a cactus-ridden desert, revealing it to be an Easter egg in the form of a platform game.[229][230][231][232] The game itself is an infinite runner, and there is no time limit in the game as it progresses faster and periodically tints to a black background. A school Chromebook administrator can disable the game.[233]

Google Chrome running on an Android phone

A beta version for Android 4.0 devices was launched on February 7, 2012, available for a limited number of countries from Google Play.[254][255]

Notable features: synchronization with desktop Chrome to provide the same bookmarks and view the same browser tabs,[256] page pre-rendering,[257] hardware acceleration.[258]

Many of the latest HTML5 features: almost all of the Web Platform's features: GPU-accelerated canvas, including CSS 3D Transforms, CSS animations, SVG, WebSocket (including binary messages), Dedicated Workers; it has overflow scroll support, strong HTML5 video support, and new capabilities such as IndexedDB, WebWorkers, Application Cache and the File APIs, date- and time-pickers, parts of the Media Capture API.[257][259] Also supports mobile oriented features such as Device Orientation and Geolocation.[259]

Mobile customizations: swipe gesture tab switching,[256] link preview allows zooming in on (multiple) links to ensure the desired one is clicked,[256] font size boosting to ensure readability regardless of the zoom level.[259]

Features missing in the mobile version include sandboxed tabs,[257] Safe Browsing,[257] apps or extensions,[258] Adobe Flash (now and in the future),[258] Native Client,[258] and the ability to export user data such a list of their opened tabs or their browsing history into portable local files.[260]

Development changes: remote debugging,[257][261] part of the browser layer has been implemented in Java, communicating with the rest of the Chromium and WebKit code through Java Native Bindings.[259] The code of Chrome for Android is a fork of the Chromium project. It is a priority to upstream most new and modified code to Chromium and WebKit to resolve the fork.[259]

The April 17, 2012, update included availability in 31 additional languages and in all countries where Google Play is available. A desktop version of a website can also be requested as opposed to a mobile version. In addition, Android users can now add bookmarks to their Android home screens if they choose and decide which apps should handle links opened in Chrome.[262]

On June 27, 2012, Google Chrome for Android exited beta and became stable.[263][264]

Chrome 18.0.1026311, released on September 26, 2012, was the first version of Chrome for Android to support mobile devices based on Intel x86.[265]

Starting from version 25, the Chrome version for Android is aligned with the desktop version, and usually new stable releases are available at the same time between the Android and the desktop version. Google released a separate Chrome for Android beta channel on January 10, 2013, with version 25.[214] As of 2013 a separate beta version of Chrome is available in the Google Play Store – it can run side by side with the stable release.[266]

Chrome is available on Apple's mobile iOS operating system as Google Chrome for iOS. Released in the Apple App Store on June 26, 2012, it supports the iPad, iPhone, and iPod Touch, and the current version requires that the device has iOS 15.0 or greater installed.[267] In accordance with Apple's requirements for browsers released through their App Store, this version of Chrome uses the iOS WebKit – which is Apple's own mobile rendering engine and components, developed for their Safari browser – therefore it is restricted from using Google's own V8 JavaScript engine.[268][269] Chrome is the default web browser for the iOS Gmail application.

In a review by Chitika, Chrome was noted as having 1.5% of the iOS web browser market as of July 18, 2012.[270] In October 2013, Chrome had 3% of the iOS browser market.[271]

On Linux distributions, support for 32-bit Intel processors ended in March 2016 although Chromium is still supported.[272] As of Chrome version 26, Linux installations of the browser may be updated only on systems that support GCC v4.6 and GTK v2.24 or later. Thus deprecated systems include (for example) Debian 6's 2.20, and RHEL 6's 2.18.[273]

Support for Google Chrome on Windows XP and Windows Vista ended in April 2016.[274] The last release of Google Chrome that can be run on Windows XP and Windows Vista was version 49.0.2623.112,[275] released on April 7, 2016,[276] then re-released on April 11, 2016.[277]

Support for Google Chrome on Windows 7 was supposed to end upon the announcement on July 15, 2021,[278] and suddenly moved to January 15, 2022, however due to the ongoing COVID-19 pandemic and enterprises are taking more time to migrate to Windows 10 or 11, the end of support date was pushed at least until January 15, 2023.[279] Support for Windows 7, Windows 8, and Windows 8.1 ended in January 2023 and the last version on Windows 7 is Chrome 109.[280][281][282]

"Windows 8 mode" was introduced in 2012 and has since been discontinued. It was provided to the developer channel, which enabled Windows 8 and Windows 8.1 users to run Chrome with a full-screen, tablet-optimized interface, with access to snapping, sharing, and search functionalities.[283] In October 2013, Windows 8 mode on the developer channel changed to use a desktop environment mimicking the interface of ChromeOS with a dedicated windowing system and taskbar for web apps.[284] This was removed on version 49 and users that have upgraded to Windows 10 will lose this feature.[285]

Google dropped support for Mac OS X Leopard with the release of Chrome 22.[286] Support for 32-bit versions of Chrome ended in November 2014 with the release of Chrome 39.[287][288][235] Support for Mac OS X Snow Leopard, Mac OS X Lion, and OS X Mountain Lion ended in April 2016 with the release of Chrome 50. Support for OS X Mavericks ended in April 2018 with the release of Chrome 66. Support for OS X Yosemite ended in January 2021 with the release of Chrome 88. Support for OS X El Capitan and macOS Sierra ended in August 2022 with the release of Chrome 104. Support for macOS High Sierra and macOS Mojave ended in September 2023 with the release of Chrome 117.[289]

Google Chrome is the basis of Google's ChromeOS operating system that ships on specific hardware from Google's manufacturing partners.[290] The user interface has a minimalist design resembling the Google Chrome browser. ChromeOS is aimed at users who spend most of their computer time on the Web; the only applications on the devices are a browser incorporating a media player and a file manager.[291][292][293][294][295]

Google announced ChromeOS on July 7, 2009.[296]

Usage share of web browsers according to StatCounter[308]

Chrome overtook Firefox in November 2011, in worldwide usage. As of September 2022, according to StatCounter, Google Chrome had 67% worldwide desktop usage share, making it the most widely used web browser.[309]

Along with Safari and Mozilla Firefox, Chrome receives a weekend "bump", which boosts its market share by as much as three percentage points on weekends, at the expense of Internet Explorer.[310]

It was reported by StatCounter, a web analytics company, that for the single day of Sunday, March 18, 2012, Chrome was the most used web browser in the world for the first time. Chrome secured 32.7% of the global web browsing on that day, while Internet Explorer followed closely behind with 32.5%.[311]

From May 14–21, 2012, Google Chrome was for the first time responsible for more Internet traffic than Microsoft's Internet Explorer, which long had held its spot as the most used web browser in the world.[312] According to StatCounter, 31.88% of web traffic was generated by Chrome for a sustained period of one week and 31.47% by Internet Explorer. Though Chrome had topped Internet Explorer for a single day's usage in the past, this was the first time it had led for one full week.[313]

At the 2012 Google I/O developers' conference, Google claimed that there were 310 million active users of Chrome, almost double the number in 2011, which was stated as 160 million active users.[314]

In June 2013, according to StatCounter, Chrome overtook Internet Explorer for the first time in the US.[315]

In August 2013, Chrome was used by 43% of internet users worldwide. This study was done by Statista, which also noted that in North America, 36% of people use Chrome, the lowest in the world.[316]

In December 2010, Google announced that to make it easier for businesses to use Chrome they would provide an official Chrome MSI package. For business use it is helpful to have full-fledged MSI packages that can be customized via transform files (.mst) – but the MSI provided with Chrome is only a very limited MSI wrapper fitted around the normal installer, and many businesses find that this arrangement does not meet their needs.[318] The normal downloaded Chrome installer puts the browser in the user's local app data directory and provides invisible background updates, but the MSI package will allow installation at the system level, providing system administrators control over the update process[319] – it was formerly possible only when Chrome was installed using Google Pack. Google also created group policy objects to fine-tune the behavior of Chrome in the business environment, for example by setting automatic updates intervals, disabling auto-updates, and configuring a home page.[320] Until version 24 the software is known not to be ready for enterprise deployments with roaming profiles or Terminal Server/Citrix environments.[321]

In 2010, Google first started supporting Chrome in enterprise environments by providing an MSI wrapper around the Chrome installer. Google starting providing group policy objects, with more added each release,[322] and today there are more than 500 policies available to control Chrome's behavior in enterprise environments.[323] In 2016, Google launched Chrome Browser Enterprise Support, a paid service enabling IT admins access to Google experts to support their browser deployment.[324] In 2019, Google launched Chrome Browser Cloud Management, a dashboard that gives business IT managers the ability to control content accessibility, app usage and browser extensions installed on its deployed computers.[325]

In September 2008, Google released a large portion of Chrome's source code as an open-source project called Chromium. This move enabled third-party developers to study the underlying source code and to help port the browser to the macOS and Linux operating systems. The Google-authored portion of Chromium is released under the permissive BSD license.[326] Other portions of the source code are subject to a variety of open-source licenses.[327] Chromium is similar to Chrome, but lacks built-in automatic updates and a built-in Flash player, as well as Google branding and has a blue-colored logo instead of the multicolored Google logo.[328][329] Chromium does not implement user RLZ tracking.[177][172][330] Initially, the Google Chrome PDF viewer, PDFium, was excluded from Chromium, but was later made open-source in May 2014.[331][332] PDFium can be used to fill PDF forms.[333]