GraphQL Benchmarks

Explore and compare the performance of the fastest GraphQL frameworks through our comprehensive benchmarks.

• Introduction
• Quick Start
• Benchmark Results
  • Throughput (Higher is better)
  • Latency (Lower is better)
• Architecture
  • WRK
  • GraphQL
  • Nginx
  • Jsonplaceholder
• GraphQL Schema
• Contribute

Introduction

This document presents a comparative analysis of several renowned GraphQL frameworks. Dive deep into the performance metrics, and get insights into their throughput and latency.

NOTE: This is a work in progress suite of benchmarks, and we would appreciate help from the community to add more frameworks or tune the existing ones for better performance.

Quick Start

Get started with the benchmarks:

1. Click on this link to set up on GitHub Codespaces.
2. Once set up in Codespaces, initiate the benchmark tests:

./setup.sh
./run_benchmarks.sh

Benchmark Results

Query	Server	Requests/sec	Latency (ms)	Relative
1	{ posts { id userId title user { id name email }}}
	Tailcall	9,823.91	10.17	82.44x
	GraphQL JIT	1,061.33	93.71	8.91x
	async-graphql	953.88	104.03	8.01x
	Caliban	764.07	131.11	6.41x
	Gqlgen	389.70	253.14	3.27x
	Netflix DGS	185.38	516.38	1.56x
	Apollo GraphQL	125.96	722.49	1.06x
	Hasura	119.16	753.08	1.00x
2	{ posts { title }}
	Tailcall	19,676.20	5.08	48.01x
	async-graphql	5,147.25	19.50	12.56x
	Caliban	4,827.56	21.25	11.78x
	Gqlgen	1,117.91	97.71	2.73x
	GraphQL JIT	1,101.18	90.63	2.69x
	Apollo GraphQL	859.04	116.89	2.10x
	Netflix DGS	803.61	125.07	1.96x
	Hasura	409.86	248.85	1.00x
3	{ greet }
	Caliban	33,062.60	3.06	23.48x
	Tailcall	26,058.70	3.87	18.50x
	Gqlgen	24,242.70	8.55	17.21x
	async-graphql	23,323.70	4.32	16.56x
	GraphQL JIT	4,455.53	22.39	3.16x
	Netflix DGS	4,149.48	28.55	2.95x
	Apollo GraphQL	3,998.37	27.43	2.84x
	Hasura	1,408.40	73.41	1.00x

1. {posts {title body user {name}}}

Throughput (Higher is better)

Latency (Lower is better)

2. {posts {title body}}

Throughput (Higher is better)

Latency (Lower is better)

3. {greet}

Throughput (Higher is better)

Latency (Lower is better)

Architecture

A client (wrk) sends requests to a GraphQL server to fetch post titles. The GraphQL server, in turn, retrieves data from an external source, jsonplaceholder.typicode.com, routed through the nginx reverse proxy.

WRK

wrk serves as our test client, sending GraphQL requests at a high rate.

GraphQL

Our tested GraphQL server. We evaluated various implementations, ensuring no caching on the GraphQL server side.

Nginx

A reverse-proxy that caches every response, mitigating rate-limiting and reducing network uncertainties.

Jsonplaceholder

The primary upstream service forming the base for our GraphQL API. We query its /posts API via the GraphQL server.

GraphQL Schema

Inspect the generated GraphQL schema employed for the benchmarks:

schema {
  query: Query
}

type Query {
  posts: [Post]
}

type Post {
  id: Int!
  userId: Int!
  title: String!
  body: String!
  user: User
}

type User {
  id: Int!
  name: String!
  username: String!
  email: String!
  phone: String
  website: String
}

Contribute

Your insights are invaluable! Test these benchmarks, share feedback, or contribute by adding more GraphQL frameworks or refining existing ones. Open an issue or a pull request, and let's build a robust benchmarking resource together!