hyper

Engine · Rust

Low-level HTTP library for Rust, built on tokio with multi-threaded async runtime.

Composite rank

Each profile of a family is worth 100 to the entry that leads it, and the composite is the sum over the family, less 2.5% for each of routing, middleware, request and response the entry does not do for you - its completeness factor, which the WebSocket and gRPC families do not carry. The field is this entry's own league: engines and reverse proxies are scored apart from frameworks. How it works.

Family	Rank	Composite
HTTP/1.1	15 of 26	3345
HTTP/2	6 of 7	664

Every result

Alpha Round, 9 runs. Requests per second is the best of three; latency, CPU and memory come from that run.

Category	Profile	Conns	Req/sec	Avg	p99	CPU	Memory
Connection	Baseline	4,096	3,003,360	1.36ms	2.68ms	6604.7%	112.4MiB
Connection	Pipelined	4,096	16,218,012	4.03ms	5.31ms	6408.6%	126.2MiB
Connection	Short-lived	4,096	2,362,875	1.70ms	4.22ms	6243.0%	189.2MiB
Efficiency	Latency-10K	1,024	9,982	72.6us	104.0us	25.1%	50MiB
Efficiency	Latency-1M	1,024	997,989	101.9us	173.0us	2432.5%	46MiB
HTTP/2	Baseline	256	6,974,787	4.24ms	4.24ms	6317.6%	156.0MiB
HTTP/2	Baseline	1,024	5,157,286	17.75ms	17.75ms	6584.6%	475.9MiB
HTTP/2	Static	256	407,071	61.68ms	61.68ms	4838.5%	268.4MiB
HTTP/2	Static	1,024	394,013	228.65ms	228.65ms	4202.2%	719.8MiB

Compared with

The Rust entries around hyper on the HTTP/1.1 composite, and the one leading them. The last column shows every profile side by side.

Entry	Composite	vs hyper
actix	6880	+106%	head to head
axum	4494	+34%	head to head
tokio-uring	3462	+3%	head to head
trillium	3261	-3%	head to head
trillium-tuned	3197	-4%	head to head

Compare hyper with any Rust entry · Compare hyper with any entry, in any language · Every Rust entry, ranked

Results updated 2026-09-07 · How the benchmark is run · The machine · Add or fix an entry