Repository files navigation

A light-weight module that bringsFetch API to Node.js.

Consider supporting us on our Open Collective:

You might be looking for thev2 docs

• Motivation
• Features
• Difference from client-side fetch
• Installation
• Loading and configuring the module
• Upgrading
• Common Usage
  • Plain text or HTML
  • JSON
  • Simple Post
  • Post with JSON
  • Post with form parameters
  • Handling exceptions
  • Handling client and server errors
  • Handling cookies
• Advanced Usage
  • Streams
  • Accessing Headers and other Metadata
  • Extract Set-Cookie Header
  • Post data using a file
  • Request cancellation with AbortSignal
• API
  • fetch(url[, options])
  • Options
    • Default Headers
    • Custom Agent
    • Custom highWaterMark
    • Insecure HTTP Parser
  • Class: Request
    • new Request(input[, options])
  • Class: Response
    • new Response([body[, options]])
    • response.ok
    • response.redirected
    • response.type
  • Class: Headers
    • new Headers([init])
  • Interface: Body
    • body.body
    • body.bodyUsed
    • body.arrayBuffer()
    • body.blob()
    • body.formData()
    • body.json()
    • body.text()
  • Class: FetchError
  • Class: AbortError
• TypeScript
• Acknowledgement
• Team-Former
• License

Instead of implementingXMLHttpRequest in Node.js to run browser-specificFetch polyfill, why not go from nativehttp tofetch API directly? Hence,node-fetch, minimal code for awindow.fetch compatible API on Node.js runtime.

See Jason Miller'sisomorphic-unfetch or Leonardo Quixada'scross-fetch for isomorphic usage (exportsnode-fetch for server-side,whatwg-fetch for client-side).

• Stay consistent withwindow.fetch API.
• Make conscious trade-off when followingWHATWG fetch spec andstream spec implementation details, document known differences.
• Use native promise and async functions.
• Use native Node streams for body, on both request and response.
• Decode content encoding (gzip/deflate/brotli) properly, and convert string output (such asres.text() andres.json()) to UTF-8 automatically.
• Useful extensions such as redirect limit, response size limit,explicit errors for troubleshooting.

• See known differences:
  • As of v3.x
  • As of v2.x
• If you happen to use a missing feature thatwindow.fetch offers, feel free to open an issue.
• Pull requests are welcomed too!

Current stable release (3.x) requires at least Node.js 12.20.0.

npm install node-fetch

importfetchfrom'node-fetch';

node-fetch from v3 is an ESM-only module - you are not able to import it withrequire().

If you cannot switch to ESM, please use v2 which remains compatible with CommonJS. Critical bug fixes will continue to be published for v2.

npm install node-fetch@2

Alternatively, you can use the asyncimport() function from CommonJS to loadnode-fetch asynchronously:

// mod.cjsconstfetch=(...args)=>import('node-fetch').then(({default:fetch})=>fetch(...args));

To usefetch() without importing it, you can patch theglobal object in node:

// fetch-polyfill.jsimportfetch,{Blob,blobFrom,blobFromSync,File,fileFrom,fileFromSync,FormData,Headers,Request,Response,}from'node-fetch'if(!globalThis.fetch){globalThis.fetch=fetchglobalThis.Headers=HeadersglobalThis.Request=RequestglobalThis.Response=Response}// index.jsimport'./fetch-polyfill'// ...

Using an old version of node-fetch? Check out the following files:

• 2.x to 3.x upgrade guide
• 1.x to 2.x upgrade guide
• Changelog

NOTE: The documentation below is up-to-date with3.x releases, if you are using an older version, please check how toupgrade.

importfetchfrom'node-fetch';constresponse=awaitfetch('https://github.com/');constbody=awaitresponse.text();console.log(body);

importfetchfrom'node-fetch';constresponse=awaitfetch('https://api.github.com/users/github');constdata=awaitresponse.json();console.log(data);

importfetchfrom'node-fetch';constresponse=awaitfetch('https://httpbin.org/post',{method:'POST',body:'a=1'});constdata=awaitresponse.json();console.log(data);

importfetchfrom'node-fetch';constbody={a:1};constresponse=awaitfetch('https://httpbin.org/post',{method:'post',body:JSON.stringify(body),headers:{'Content-Type':'application/json'}});constdata=awaitresponse.json();console.log(data);

URLSearchParams is available on the global object in Node.js as of v10.0.0. Seeofficial documentation for more usage methods.

NOTE: TheContent-Type header is only set automatically tox-www-form-urlencoded when an instance ofURLSearchParams is given as such:

importfetchfrom'node-fetch';constparams=newURLSearchParams();params.append('a',1);constresponse=awaitfetch('https://httpbin.org/post',{method:'POST',body:params});constdata=awaitresponse.json();console.log(data);

NOTE: 3xx-5xx responses areNOT exceptions, and should be handled inthen(), see the next section.

Wrapping the fetch function into atry/catch block will catchall exceptions, such as errors originating from node core libraries, like network errors, and operational errors which are instances of FetchError. See theerror handling document for more details.

importfetchfrom'node-fetch';try{awaitfetch('https://domain.invalid/');}catch(error){console.log(error);}

It is common to create a helper function to check that the response contains no client (4xx) or server (5xx) error responses:

importfetchfrom'node-fetch';classHTTPResponseErrorextendsError{constructor(response){super(`HTTP Error Response:${response.status}${response.statusText}`);this.response=response;}}constcheckStatus=response=>{if(response.ok){// response.status >= 200 && response.status < 300returnresponse;}else{thrownewHTTPResponseError(response);}}constresponse=awaitfetch('https://httpbin.org/status/400');try{checkStatus(response);}catch(error){console.error(error);consterrorBody=awaiterror.response.text();console.error(`Error body:${errorBody}`);}

Cookies are not stored by default. However, cookies can be extracted and passed by manipulating request and response headers. SeeExtract Set-Cookie Header for details.

The "Node.js way" is to use streams when possible. You can piperes.body to another stream. This example usesstream.pipeline to attach stream error handlers and wait for the download to complete.

import{createWriteStream}from'node:fs';import{pipeline}from'node:stream';import{promisify}from'node:util'importfetchfrom'node-fetch';conststreamPipeline=promisify(pipeline);constresponse=awaitfetch('https://github.githubassets.com/images/modules/logos_page/Octocat.png');if(!response.ok)thrownewError(`unexpected response${response.statusText}`);awaitstreamPipeline(response.body,createWriteStream('./octocat.png'));

In Node.js 14 you can also use async iterators to readbody; however, be careful to catcherrors -- the longer a response runs, the more likely it is to encounter an error.

importfetchfrom'node-fetch';constresponse=awaitfetch('https://httpbin.org/stream/3');try{forawait(constchunkofresponse.body){console.dir(JSON.parse(chunk.toString()));}}catch(err){console.error(err.stack);}

In Node.js 12 you can also use async iterators to readbody; however, async iterators with streamsdid not mature until Node.js 14, so you need to do some extra work to ensure you handle errorsdirectly from the stream and wait on it response to fully close.

importfetchfrom'node-fetch';constread=asyncbody=>{leterror;body.on('error',err=>{error=err;});forawait(constchunkofbody){console.dir(JSON.parse(chunk.toString()));}returnnewPromise((resolve,reject)=>{body.on('close',()=>{error ?reject(error) :resolve();});});};try{constresponse=awaitfetch('https://httpbin.org/stream/3');awaitread(response.body);}catch(err){console.error(err.stack);}

importfetchfrom'node-fetch';constresponse=awaitfetch('https://github.com/');console.log(response.ok);console.log(response.status);console.log(response.statusText);console.log(response.headers.raw());console.log(response.headers.get('content-type'));

Unlike browsers, you can access rawSet-Cookie headers manually usingHeaders.raw(). This is anode-fetch only API.

importfetchfrom'node-fetch';constresponse=awaitfetch('https://example.com');// Returns an array of values, instead of a string of comma-separated valuesconsole.log(response.headers.raw()['set-cookie']);

importfetch,{Blob,blobFrom,blobFromSync,File,fileFrom,fileFromSync,}from'node-fetch'constmimetype='text/plain'constblob=fileFromSync('./input.txt',mimetype)consturl='https://httpbin.org/post'constresponse=awaitfetch(url,{method:'POST',body:blob})constdata=awaitresponse.json()console.log(data)

node-fetch comes with a spec-compliantFormData implementations for postingmultipart/form-data payloads

importfetch,{FormData,File,fileFrom}from'node-fetch'consthttpbin='https://httpbin.org/post'constformData=newFormData()constbinary=newUint8Array([97,98,99])constabc=newFile([binary],'abc.txt',{type:'text/plain'})formData.set('greeting','Hello, world!')formData.set('file-upload',abc,'new name.txt')constresponse=awaitfetch(httpbin,{method:'POST',body:formData})constdata=awaitresponse.json()console.log(data)

If you for some reason need to post a stream coming from any arbitrary place,then you can append aBlob or aFile look-a-like item.

The minimum requirement is that it has:

1. ASymbol.toStringTag getter or property that is eitherBlob orFile
2. A known size.
3. And either astream() method or aarrayBuffer() method that returns a ArrayBuffer.

Thestream() must return any async iterable object as long as it yields Uint8Array (or Buffer)so Node.Readable streams and whatwg streams works just fine.

formData.append('upload',{[Symbol.toStringTag]:'Blob',size:3,*stream(){yieldnewUint8Array([97,98,99])},arrayBuffer(){returnnewUint8Array([97,98,99]).buffer}},'abc.txt')

You may cancel requests withAbortController. A suggested implementation isabort-controller.

An example of timing out a request after 150ms could be achieved as the following:

importfetch,{AbortError}from'node-fetch';// AbortController was added in node v14.17.0 globallyconstAbortController=globalThis.AbortController||awaitimport('abort-controller')constcontroller=newAbortController();consttimeout=setTimeout(()=>{controller.abort();},150);try{constresponse=awaitfetch('https://example.com',{signal:controller.signal});constdata=awaitresponse.json();}catch(error){if(errorinstanceofAbortError){console.log('request was aborted');}}finally{clearTimeout(timeout);}

Seetest cases for more examples.

• url A string representing the URL for fetching
• optionsOptions for the HTTP(S) request
• Returns:Promise<Response>

Perform an HTTP(S) fetch.

url should be an absolute URL, such ashttps://example.com/. A path-relative URL (/file/under/root) or protocol-relative URL (//can-be-http-or-https.com/) will result in a rejectedPromise.

The default values are shown after each option key.

{// These properties are part of the Fetch Standardmethod:'GET',headers:{},// Request headers. format is the identical to that accepted by the Headers constructor (see below)body:null,// Request body. can be null, or a Node.js Readable streamredirect:'follow',// Set to `manual` to extract redirect headers, `error` to reject redirectsignal:null,// Pass an instance of AbortSignal to optionally abort requests// The following properties are node-fetch extensionsfollow:20,// maximum redirect count. 0 to not follow redirectcompress:true,// support gzip/deflate content encoding. false to disablesize:0,// maximum response body size in bytes. 0 to disableagent:null,// http(s).Agent instance or function that returns an instance (see below)highWaterMark:16384,// the maximum number of bytes to store in the internal buffer before ceasing to read from the underlying resource.insecureHTTPParser:false// Use an insecure HTTP parser that accepts invalid HTTP headers when `true`.}

If no values are set, the following request headers will be sent automatically:

Note: whenbody is aStream,Content-Length is not set automatically.

Theagent option allows you to specify networking related options which are out of the scope of Fetch, including and not limited to the following:

• Support self-signed certificate
• Use only IPv4 or IPv6
• Custom DNS Lookup

Seehttp.Agent for more information.

If no agent is specified, the default agent provided by Node.js is used. Note thatthis changed in Node.js 19 to havekeepalive true by default. If you wish to enablekeepalive in an earlier version of Node.js, you can override the agent as per the following code sample.

In addition, theagent option accepts a function that returnshttp(s).Agent instance given currentURL, this is useful during a redirection chain across HTTP and HTTPS protocol.

importhttpfrom'node:http';importhttpsfrom'node:https';consthttpAgent=newhttp.Agent({keepAlive:true});consthttpsAgent=newhttps.Agent({keepAlive:true});constoptions={agent:function(_parsedURL){if(_parsedURL.protocol=='http:'){returnhttpAgent;}else{returnhttpsAgent;}}};

Stream on Node.js have a smaller internal buffer size (16kB, akahighWaterMark) from client-side browsers (>1MB, not consistent across browsers). Because of that, when you are writing an isomorphic app and usingres.clone(), it will hang with large response in Node.

The recommended way to fix this problem is to resolve cloned response in parallel:

importfetchfrom'node-fetch';constresponse=awaitfetch('https://example.com');constr1=response.clone();constresults=awaitPromise.all([response.json(),r1.text()]);console.log(results[0]);console.log(results[1]);

If for some reason you don't like the solution above, since3.x you are able to modify thehighWaterMark option:

importfetchfrom'node-fetch';constresponse=awaitfetch('https://example.com',{// About 1MBhighWaterMark:1024*1024});constresult=awaitres.clone().arrayBuffer();console.dir(result);

Passed through to theinsecureHTTPParser option on http(s).request. Seehttp.request for more information.

Theredirect: 'manual' option for node-fetch is different from the browser & specification, whichresults in anopaque-redirect filtered response.node-fetch gives you the typicalbasic filtered response instead.

importfetchfrom'node-fetch';constresponse=awaitfetch('https://httpbin.org/status/301',{redirect:'manual'});if(response.status===301||response.status===302){constlocationURL=newURL(response.headers.get('location'),response.url);constresponse2=awaitfetch(locationURL,{redirect:'manual'});console.dir(response2);}

An HTTP(S) request containing information about URL, method, headers, and the body. This class implements theBody interface.

Due to the nature of Node.js, the following properties are not implemented at this moment:

• type
• destination
• mode
• credentials
• cache
• integrity
• keepalive

The following node-fetch extension properties are provided:

• follow
• compress
• counter
• agent
• highWaterMark

Seeoptions for exact meaning of these extensions.

(spec-compliant)

• input A string representing a URL, or anotherRequest (which will be cloned)
• optionsOptions for the HTTP(S) request

Constructs a newRequest object. The constructor is identical to that in thebrowser.

In most cases, directlyfetch(url, options) is simpler than creating aRequest object.

An HTTP(S) response. This class implements theBody interface.

The following properties are not implemented in node-fetch at this moment:

• trailer

(spec-compliant)

• body AString orReadable stream
• options AResponseInit options dictionary

Constructs a newResponse object. The constructor is identical to that in thebrowser.

Because Node.js does not implement service workers (for which this class was designed), one rarely has to construct aResponse directly.

(spec-compliant)

Convenience property representing if the request ended normally. Will evaluate to true if the response status was greater than or equal to 200 but smaller than 300.

(spec-compliant)

Convenience property representing if the request has been redirected at least once. Will evaluate to true if the internal redirect counter is greater than 0.

(deviation from spec)

Convenience property representing the response's type. node-fetch only supports'default' and'error' and does not make use offiltered responses.

This class allows manipulating and iterating over a set of HTTP headers. All methods specified in theFetch Standard are implemented.

(spec-compliant)

• init Optional argument to pre-fill theHeaders object

Construct a newHeaders object.init can be eithernull, aHeaders object, an key-value map object or any iterable object.

// Example adapted from https://fetch.spec.whatwg.org/#example-headers-classimport{Headers}from'node-fetch';constmeta={'Content-Type':'text/xml'};constheaders=newHeaders(meta);// The above is equivalent toconstmeta=[['Content-Type','text/xml']];constheaders=newHeaders(meta);// You can in fact use any iterable objects, like a Map or even another Headersconstmeta=newMap();meta.set('Content-Type','text/xml');constheaders=newHeaders(meta);constcopyOfHeaders=newHeaders(headers);

Body is an abstract interface with methods that are applicable to bothRequest andResponse classes.

(deviation from spec)

• Node.jsReadable stream

Data are encapsulated in theBody object. Note that while theFetch Standard requires the property to always be a WHATWGReadableStream, in node-fetch it is a Node.jsReadable stream.

(spec-compliant)

• Boolean

A boolean property for if this body has been consumed. Per the specs, a consumed body cannot be used again.

fetch comes with methods to parsemultipart/form-data payloads as well asx-www-form-urlencoded bodies using.formData() this comes from the idea thatService Worker can intercept such messages before it's sent to the server toalter them. This is useful for anybody building a server so you can use it toparse & consume payloads.

importhttpfrom'node:http'import{Response}from'node-fetch'http.createServer(asyncfunction(req,res){constformData=awaitnewResponse(req,{headers:req.headers// Pass along the boundary value}).formData()constallFields=[...formData]constfile=formData.get('uploaded-files')constarrayBuffer=awaitfile.arrayBuffer()consttext=awaitfile.text()constwhatwgReadableStream=file.stream()// other was to consume the request could be to do:constjson=awaitnewResponse(req).json()consttext=awaitnewResponse(req).text()constarrayBuffer=awaitnewResponse(req).arrayBuffer()constblob=awaitnewResponse(req,{headers:req.headers// So that `type` inherits `Content-Type`}.blob()})

(node-fetch extension)

An operational error in the fetching process. SeeERROR-HANDLING.md for more info.

(node-fetch extension)

An Error thrown when the request is aborted in response to anAbortSignal'sabort event. It has aname property ofAbortError. SeeERROR-HANDLING.MD for more info.

Since3.x types are bundled withnode-fetch, so you don't need to install any additional packages.

For older versions please use the type definitions fromDefinitelyTyped:

npm install --save-dev @types/node-fetch@2.x

Thanks togithub/fetch for providing a solid implementation reference.

David Frank	Jimmy Wärting	Antoni Kepinski	Richie Bendall	Gregor Martynus

• Timothy Gu
• Jared Kantrowitz

MIT