HTTP/2 is the second major version of the Hypertext Transfer Protocol, the foundational communication standard that governs how data is exchanged between web browsers and servers. Standardized in 2015 by the Internet Engineering Task Force (IETF), it was designed to address the performance limitations of its predecessor, HTTP/1.1, which had remained largely unchanged since 1997.
To understand why HTTP/2 matters, it helps to know how HTTP/1.1 handles requests. Under the older protocol, a browser typically opens multiple separate connections to a server and sends requests one at a time per connection — a model that creates significant inefficiency as modern web pages load dozens or even hundreds of individual resources such as scripts, stylesheets, and images. HTTP/2 resolves this through a mechanism called multiplexing, which allows multiple requests and responses to travel simultaneously over a single connection. This eliminates the queuing bottleneck known as head-of-line blocking, where a slow or large resource delays everything behind it.
Another major improvement is header compression. Every HTTP request carries metadata in the form of headers — information such as cookies, content types, and authentication tokens. In HTTP/1.1, these headers are sent as plain text with every request, even when they are largely identical between calls. HTTP/2 introduces a compression algorithm called HPACK that significantly reduces the size of this repeated overhead, lowering the amount of data transferred and reducing latency on each round trip.
HTTP/2 also introduces server push, a feature that allows a server to proactively send resources to the browser before they are explicitly requested. If a server knows that a particular stylesheet or script will be needed to render a page, it can begin transmitting that asset alongside the initial HTML response, reducing the number of round trips required to fully load a page. In practice, server push requires careful configuration to be effective, and its adoption has been uneven across web hosting environments.
From a web performance perspective, HTTP/2 can meaningfully reduce page load times, particularly on connections with higher latency such as mobile networks. The gains are most pronounced on pages with many small resources, where the multiplexing advantage is greatest. Enabling HTTP/2 typically requires no changes to a website's code — it is configured at the server or hosting level, and modern browsers support it automatically when the server advertises the protocol over a secure HTTPS connection.
It is worth noting that HTTP/2 has itself been followed by HTTP/3, which replaces the underlying TCP transport layer with a protocol called QUIC to further reduce latency and improve performance on unreliable connections. HTTP/2, however, remains widely deployed and represents a significant baseline improvement over HTTP/1.1 for the vast majority of web infrastructure in use today.