Hyperttp ⚡

HTTP without boilerplate.

English | Русский

An HTTP client for Node.js, Bun, Deno, and the Browser that provides retries, caching, rate limiting, metrics, parsing, serialization, and interceptors out of the box.

import { HyperClient } from "hyperttp";

const client = new HyperClient();

// Automatically parses response based on the second argument
const user = await client.get("https://api.example.com/users/1", "json");

console.log(user);

No manual setup. No unnecessary code. Just make requests.

What is Hyperttp?

Most HTTP clients solve one problem:

"Send an HTTP request."

But real applications need much more:

• retries
• rate limiting
• caching
• request deduplication
• parsing
• serialization
• metrics
• interceptors
• request queues
• timeouts

With most libraries, this turns into installing many additional packages and building your own wrappers.

Hyperttp provides these capabilities through a single client.

Why Hyperttp?

Instead of building your own networking layer:

fetch
 ├── retry
 ├── cache
 ├── parser
 ├── metrics
 ├── serializer
 ├── queue
 ├── timeout
 └── custom wrappers

Hyperttp provides a ready-to-use infrastructure:

Hyperttp
 ├── Retry
 ├── Cache
 ├── Metrics
 ├── Parser
 ├── Serializer
 ├── Queue
 ├── Interceptors
 ├── Timeout
 └── Rate Limiter

Create a client.

Start making requests.

Quick Start

1. Install core package

# Node.js
npm install hyperttp
pnpm add hyperttp

# Bun
bun add hyperttp

# Deno
deno add npm:hyperttp

2. Basic Usage

import { HyperClient } from "hyperttp";

const client = new HyperClient({
  baseURL: "https://api.example.com",
});

await client.get("/users", "json");
await client.post("/users", { name: "John" }, "json");
await client.put("/users/1", { name: "Alice" }, "json");
await client.delete("/users/1");

Hyperttp vs Other HTTP Clients

| Feature | fetch | Axios | Ky | Hyperttp | | ------------------------ | ----- | ------ | ------- | -------- | | Promise API | ✅ | ✅ | ✅ | ✅ | | Retry | ❌ | Plugin | ✅ | ✅ | | Cache | ❌ | ❌ | ❌ | ✅ | | Metrics | ❌ | ❌ | ❌ | ✅ | | Rate Limiter | ❌ | ❌ | ❌ | ✅ | | Queue | ❌ | ❌ | ❌ | ✅ | | Parser | ❌ | ❌ | Partial | ✅ | | Serializer | ❌ | ❌ | ❌ | ✅ | | Interceptors | ❌ | ✅ | Hooks | ✅ | | Transport auto-detection | ❌ | ❌ | ❌ | ✅ |

Who is Hyperttp for?

Perfect for:

• REST APIs & SDKs
• CLI applications & Discord bots
• Backend & Serverless services
• Web scraping
• Enterprise applications

Advanced Configuration

const client = new HyperClient({
  baseURL: "https://api.example.com",

  retry: {
    retries: 3,
  },

  cache: {
    ttl: 60_000,
  },

  timeout: 30_000,
});

Architecture

Application
     │
     ▼
 HyperClient
     │
 Plugin Pipeline
     │
 Transport (Auto-detected)
     │
    HTTP

All components are independent and replaceable.

Runtime Transports

Hyperttp works out of the box in any environment using standard globalThis.fetch.

For maximum performance and native features, you can optionally install an optimized transport package tailored for your specific runtime:

| Runtime | Package | Description | | ------- | ---------------------------- | -------------------------------------------- | | Node.js | @hyperttp/transport-undici | Uses high-performance native undici client | | Bun | @hyperttp/transport-bun | Leverages native Bun.fetch optimizations | | Deno | @hyperttp/transport-deno | Optimized for native Deno network layer |

Once installed, no additional configuration is required:

import { HyperClient } from "hyperttp";

const client = new HyperClient();
// Automatically discovers and uses UndiciTransport on Node.js, BunTransport on Bun, etc.

Ecosystem

• hyperttp
• @hyperttp/core
• @hyperttp/types
• @hyperttp/parser
• @hyperttp/cache
• @hyperttp/retry
• @hyperttp/metrics
• @hyperttp/serializer
• @hyperttp/transport-undici
• @hyperttp/transport-bun
• @hyperttp/transport-deno

Performance

Hyperttp is designed with zero-allocation middleware pipelines and minimal overhead. By using native runtime transports, it achieves near-raw-fetch speeds while providing a full feature set.

Detailed benchmarks comparing memory footprints and request execution times across runtimes are available in the IT-IF-OR/bench repository.

License

MIT