Unlock Node.js’s Built‑in fetch(): How It Works and Why Use It

Node.js introduced the native fetch() API in its latest version. This popular cross‑platform HTTP client runs in browsers and Web/Service Workers, and as of v17.5.0 it is available as an experimental feature, potentially becoming standard in future LTS releases.

What is fetch()?

The fetch() API provides a WHATWG‑standardized interface for retrieving resources. It is a Promise‑based HTTP client that supports many advanced HTTP features while focusing on the most common scenario: sending simplified HTTP requests.

The core of the Fetch API consists of four interfaces:

fetch() – the entry point for initiating requests

Headers class – handles HTTP request/response headers

Request class – represents a request

Response class – represents a response

Here is a simple example that puts them together:

const res = await fetch('https://example/api/list');
const json = await res.json();
console.log(json);

The advantage of standardization is that the usage mirrors the Fetch API in browsers.

Why use it?

There are two main reasons to use fetch() in Node.js:

The Node.js community is debating how to evolve the HTTP stack in a way familiar to client‑side developers, moving beyond the current core HTTP model and supporting HTTP/2‑3 without adding heavy burdens; fetch() is the first step in that discussion.

The implementation is based on Undici , a fast, reliable, spec‑compliant HTTP/1.1 client built directly on sockets. Compared with existing implementations, fetch() can significantly reduce latency and increase throughput.

How to enable?

In Node.js 17.5, fetch() is experimental. Run scripts with the --experimental-fetch flag to use it. Stay tuned to the FedJavaScript channel for future updates.