HTTP/2 is a binary, multiplexed version of HTTP that uses streams, header compression (HPACK), and optional server push to reduce latency and improve performance over a single TCP connection.
HTTP 2.0
HTTP/2 is the second major version of the Hypertext Transfer Protocol, originally standardized as RFC 7540 and now maintained as RFC 9113. It keeps the same HTTP semantics developers already know — methods, status codes, and headers — but changes how those messages travel over the wire, moving from plain text to a binary framing layer that multiplexes many requests across a single TCP connection.
- Binary framing - Messages are split into binary frames rather than plain-text lines, which is more compact and less error-prone to parse.
- Multiplexing - Many requests and responses share one connection at the same time, removing the head-of-line blocking that plagued HTTP/1.1.
- Header compression (HPACK) - Repeated header fields are compressed with HPACK, cutting the overhead that dominates chatty API traffic.
- Stream prioritization and server push - Clients can hint at priority, and servers can optionally push related resources ahead of a request.
In real API operations, HTTP/2 shows up quietly underneath most modern gateways, CDNs, and cloud load balancers, where its multiplexing meaningfully lowers latency for consumers making many small calls. When I score providers, HTTP/2 support is table stakes rather than a differentiator — it is where the industry landed before the shift to HTTP/3 over QUIC, and it remains the workhorse transport for the majority of API and web traffic today.
Referenced on the API Evangelist blog
Where this standard shows up across sixteen years of my writing at apievangelist.com — how it fits into API design, governance, and the agentic turn.