RTP Benchmark: GraphQL vs REST vs Go/HTMX

High load (1200 rps) — watch the stacks diverge under pressure · Read the full analysis

Speed: 0s

read-one

Target: 1200 rps · 18000 requests · 76.2% success · 15s duration

p50 Latency

Node/GraphQL

Node/REST

Go/HTMX

p95 Latency

Node/GraphQL

Node/REST

Go/HTMX

Throughput

Node/GraphQL

Node/REST

Go/HTMX

Success Rate

Node/GraphQL

Node/REST

Go/HTMX

CPU %

Node/GraphQL

Node/REST

Go/HTMX

Memory (RSS)

Node/GraphQL

Node/REST

Go/HTMX

Stack	Rate	Success	p50	p95	Max	Throughput	CPU	RSS
Node/GraphQL	1200	76.2%	13.8s	30.0s	30.0s	305 rps	121%	820 MB
Node/REST	965	45.8%	1.8s	29.4s	30.0s	200 rps	38%	294 MB
Go/HTMX	1191	100.0%	9.1s	9.9s	9.9s	741 rps	374%	501 MB

read-many

Target: 20 rps · 300 requests · 100.0% success · 15s duration

p50 Latency

Node/GraphQL

Node/REST

Go/HTMX

p95 Latency

Node/GraphQL

Node/REST

Go/HTMX

Throughput

Node/GraphQL

Node/REST

Go/HTMX

Success Rate

Node/GraphQL

Node/REST

Go/HTMX

CPU %

Node/GraphQL

Node/REST

Go/HTMX

Memory (RSS)

Node/GraphQL

Node/REST

Go/HTMX

Stack	Rate	Success	p50	p95	Max	Throughput	CPU	RSS
Node/GraphQL	20	100.0%	6.9ms	8.1ms	9.8ms	20 rps	13%	953 MB
Node/REST	20	100.0%	3.8ms	4.7ms	10.9ms	20 rps	7%	405 MB
Go/HTMX	20	100.0%	20.7ms	23.1ms	25.2ms	20 rps	16%	96 MB

read-small

Target: 1166 rps · 17545 requests · 49.0% success · 15s duration

p50 Latency

Node/GraphQL

Node/REST

Go/HTMX

p95 Latency

Node/GraphQL

Node/REST

Go/HTMX

Throughput

Node/GraphQL

Node/REST

Go/HTMX

Success Rate

Node/GraphQL

Node/REST

Go/HTMX

CPU %

Node/GraphQL

Node/REST

Go/HTMX

Memory (RSS)

Node/GraphQL

Node/REST

Go/HTMX

Stack	Rate	Success	p50	p95	Max	Throughput	CPU	RSS
Node/GraphQL	1166	49.0%	470.6ms	15.5s	28.1s	270 rps	38%	953 MB
Node/REST	1200	100.0%	1.3ms	4.3ms	92.1ms	1200 rps	90%	367 MB
Go/HTMX	1200	100.0%	2.8ms	5.6ms	14.8ms	1200 rps	312%	94 MB