Skip to content

Instantly share code, notes, and snippets.

What would you like to do?
Encode an ArrayBuffer as a base64 string
// Converts an ArrayBuffer directly to base64, without any intermediate 'convert to string then
// use window.btoa' step. According to my tests, this appears to be a faster approach:
Copyright 2011 Jon Leighton
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.
function base64ArrayBuffer(arrayBuffer) {
var base64 = ''
var encodings = 'ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/'
var bytes = new Uint8Array(arrayBuffer)
var byteLength = bytes.byteLength
var byteRemainder = byteLength % 3
var mainLength = byteLength - byteRemainder
var a, b, c, d
var chunk
// Main loop deals with bytes in chunks of 3
for (var i = 0; i < mainLength; i = i + 3) {
// Combine the three bytes into a single integer
chunk = (bytes[i] << 16) | (bytes[i + 1] << 8) | bytes[i + 2]
// Use bitmasks to extract 6-bit segments from the triplet
a = (chunk & 16515072) >> 18 // 16515072 = (2^6 - 1) << 18
b = (chunk & 258048) >> 12 // 258048 = (2^6 - 1) << 12
c = (chunk & 4032) >> 6 // 4032 = (2^6 - 1) << 6
d = chunk & 63 // 63 = 2^6 - 1
// Convert the raw binary segments to the appropriate ASCII encoding
base64 += encodings[a] + encodings[b] + encodings[c] + encodings[d]
// Deal with the remaining bytes and padding
if (byteRemainder == 1) {
chunk = bytes[mainLength]
a = (chunk & 252) >> 2 // 252 = (2^6 - 1) << 2
// Set the 4 least significant bits to zero
b = (chunk & 3) << 4 // 3 = 2^2 - 1
base64 += encodings[a] + encodings[b] + '=='
} else if (byteRemainder == 2) {
chunk = (bytes[mainLength] << 8) | bytes[mainLength + 1]
a = (chunk & 64512) >> 10 // 64512 = (2^6 - 1) << 10
b = (chunk & 1008) >> 4 // 1008 = (2^6 - 1) << 4
// Set the 2 least significant bits to zero
c = (chunk & 15) << 2 // 15 = 2^4 - 1
base64 += encodings[a] + encodings[b] + encodings[c] + '='
return base64
Copy link

otaroo commented Sep 26, 2017

Thank you very much

Copy link

stewartsims commented Dec 14, 2017

Very handy. ES6 version works well. 👍

Copy link

Activesite commented Jan 25, 2018

Thank you very much. You save my day 👍

Copy link

proton5000 commented Feb 6, 2018

This is not work in EDGE and Internet Explorer 11. Please, help me :'(

Copy link

TFrascaroli commented Feb 7, 2018

@hookenz - I know this is quite a long way down the road (almost a year after your last coment), but I don't think you see what's going on here and I fear that other people may fall into the same trap. You see, OP is giving us a function to convert ANY arbitrary ArrayBuffer to a base64 string in javascript. Your solution works great aswell, except that it ONLY works when the underlying array of bytes comes from a UTF-8 encoded string, meaning that it can not be used to convert all byte arrays, only those that don't have "invalid" bytes in it, namely the infamous '\0' byte. So it doesn't really matter if it's faster if it doesn't work for all intended cases.

In case anyone is wondering why this is so, it's because of the way String.fromCharCode() works.
For a really small and homemade test, you can do this:

  • Assume hookenzBase64 is your function
  • Assume jonleightonBase64 is OP function
var b = new Uint16Array(25);
b.set([1,2,3,4,5], 3);

I hope this clarifies things for anyone else looking into these two functions ;)

Copy link

TFrascaroli commented Feb 7, 2018

@proton5000 - Tested on Edge/15.15063 and it works for me (with the really barebones test that I posted just now).
Also, it should work, according to CanIUse

Copy link

carpiediem commented Feb 23, 2018

I'd like to validate the output string by rendering the image again. What do I need to add to the front to set it as the src attribute of an img tag? I tried this, but just got a broken image. (Failed to load resource: net::ERR_FILE_NOT_FOUND)

<img src="data:image/png;base64,[function result]">


Copy link

eladkarako commented Mar 9, 2018


  //get ArrayBuffer from somewhere, for example:
  //FileReader's readAsArrayBuffer ".result"
  //and convert them to a BASE64-text.

function simpler_way(buffer){
  var tmp;

  tmp = (new TextDecoder("utf-8")).decode(buffer); //to UTF-8 text.
  tmp = unescape(encodeURIComponent(tmp));         //to binary-string.
  tmp = btoa(tmp);                                 //BASE64.
  return tmp;

//you can stop here.
//but if you need to return an ArrayBuffer of the BASE64 result,
//for example to be passed from a Worker back to a client, 
//by using 'postMessage' + Transferable object (much faster see:,
//uncomment lines below:
//tmp = (new TextEncoder("utf-8")).encode(tmp);    //to Uint8Array.
//tmp = tmp.buffer;                                //to ArrayBuffer.
//return tmp;

used in this repository,
as a Worker-FileReader example
(a no-upload file to Base64 converter )

Copy link

matsumeri commented Apr 24, 2018

Hi, this library exist in bower?

Copy link

dgboss commented Jun 5, 2018

Thank you! I finally came across this gist after a few hours of banging my head against a wall. I was fetching a PNG as an ArrayBuffer and needed to convert to a base64 encoding so I could automatically save the png.

Copy link

talhajunaidd commented Nov 8, 2018

I was using node's Buffer to encode values and then decode in python using base64 package and produced false results on python side.
Using this gist I was able to successfully decode values in python.
You saved my day. Many many Thanks for sharing.

Copy link

jetobe95 commented Nov 28, 2018


Copy link

Teamop commented Dec 7, 2018

based on,
a more easier way:

function bufferToBase64(buffer) {
     const binary = String.fromCharCode.apply(null, buffer);
     return window.btoa(binary);

Copy link

paschalidi commented Feb 17, 2019

@Teamop this way wont work for ArrayBuffer larger than 30k bytes.

Copy link

paschalidi commented Feb 17, 2019

I ended up using this

  "use strict";

  var chars = "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/";

  // Use a lookup table to find the index.
  var lookup = new Uint8Array(256);
  for (var i = 0; i < chars.length; i++) {
    lookup[chars.charCodeAt(i)] = i;

  exports.encode = function(arraybuffer) {
    var bytes = new Uint8Array(arraybuffer),
    i, len = bytes.length, base64 = "";

    for (i = 0; i < len; i+=3) {
      base64 += chars[bytes[i] >> 2];
      base64 += chars[((bytes[i] & 3) << 4) | (bytes[i + 1] >> 4)];
      base64 += chars[((bytes[i + 1] & 15) << 2) | (bytes[i + 2] >> 6)];
      base64 += chars[bytes[i + 2] & 63];

    if ((len % 3) === 2) {
      base64 = base64.substring(0, base64.length - 1) + "=";
    } else if (len % 3 === 1) {
      base64 = base64.substring(0, base64.length - 2) + "==";

    return base64;

  exports.decode =  function(base64) {
    var bufferLength = base64.length * 0.75,
    len = base64.length, i, p = 0,
    encoded1, encoded2, encoded3, encoded4;

    if (base64[base64.length - 1] === "=") {
      if (base64[base64.length - 2] === "=") {

    var arraybuffer = new ArrayBuffer(bufferLength),
    bytes = new Uint8Array(arraybuffer);

    for (i = 0; i < len; i+=4) {
      encoded1 = lookup[base64.charCodeAt(i)];
      encoded2 = lookup[base64.charCodeAt(i+1)];
      encoded3 = lookup[base64.charCodeAt(i+2)];
      encoded4 = lookup[base64.charCodeAt(i+3)];

      bytes[p++] = (encoded1 << 2) | (encoded2 >> 4);
      bytes[p++] = ((encoded2 & 15) << 4) | (encoded3 >> 2);
      bytes[p++] = ((encoded3 & 3) << 6) | (encoded4 & 63);

    return arraybuffer;

Copy link

tinkotvas commented May 13, 2019

function bufferToBase64(buffer) {
     const binary = String.fromCharCode.apply(null, buffer);
     return window.btoa(binary);

@Teamop this way wont work for ArrayBuffer larger than 30k bytes.

While this thread was about "Converts an ArrayBuffer directly to base64, without any intermediate 'convert to string then use window.btoa' step" so I'm going a bit off topic, but there is a way to avoid the "larger than 30k bytes" issues using reduce.

bufferToBase64(buffer) {
    return btoa(new Uint8Array(buffer).reduce((data, byte)=> {
      return data + String.fromCharCode(byte);
    }, ''));

Further example using this with angular7 for image source (say you're storing images in a database)

  bufferToBase64ImageSource(buffer) {
    const base64String = btoa(new Uint8Array(buffer).reduce((data, byte)=> {
      return data + String.fromCharCode(byte);
    }, ''));
    return this.domSanitzer.bypassSecurityTrustUrl('data:image/jpg;base64, ' + base64String);

Copy link

bolmaster2 commented Apr 5, 2020

@Teamop this way wont work for ArrayBuffer larger than 30k bytes.

Can you elaborate why it wouldn't work? I've been using it successfully in both Firefox and Chrome?

Copy link

spalt08 commented Apr 30, 2020

I've made a benchmark:

The code below shows stable and fast result among browsers.

function bufferToBase64(buffer) {
     const binary = String.fromCharCode.apply(null, buffer);
     return window.btoa(binary);

However, original approach is a little bit faster in Safari

Copy link

odbol commented Aug 28, 2020

Warning: Using @spalt08's method with the recursion will fail with a RangeError: Maximum call stack size exceeded error for large buffers.

Better to use a good ol' for loop, like @hookenz. Sometimes functional programming is terribly inefficient.

Copy link

spalt08 commented Aug 31, 2020

there is no recursion here. If your payload is big, String.fromCharCode.apply will overflow execution stack. You can use for loop for converting buffer to binary string in these cases

Copy link

hookenz commented Oct 22, 2020

@TFrascaroli - thanks for the feedback in 2016! As I recall I needed the utmost performance when converting a UTF-8 encoded string. It was something to do with encoding images for something like an MJPEG stream so performance was important and 40ms made a difference. Anyway, thanks for pointing out the differences.

Copy link

TFrascaroli commented Nov 18, 2020

@hookenz To be perfectly frank, I don't even remember what I used this for. But I remember that I needed a way to convert binary data into base64, not just binary representation of a string, but actual binary data with a bunch of '\0's in it. Anyway, if I helped in any way I'm glad I could do so 👍

Copy link

wis commented Nov 20, 2020

I found that converting to a string (with String.fromCharCode) and then base64 encoding (with btoa) is faster, it may have changed

edit: I just noticed and fixed a flaw in the test above, a new arraybuffer is being created in the function, and the whole point of this gist is avoiding copying. I fixed it here
turns out that base64 encoding an arraybuffer to string is faster than creating a DOMString then base64 encoding to a DOMstring.

edit2: with a big arraybuffer the performance tanks and performance of native String.fromCharCode stays more consistent

Copy link

peterbajwa commented Nov 22, 2020

Can anyone please help me know, is this base64 conversion specific to Javascript or is equivalent to other languages Base64 conversion methods. eg. Ruby has below code to convert the byte array to base64 string.

Copy link

TFrascaroli commented Nov 22, 2020

@peterbajwa Base64 is base64, no matter who produces it or who consumes it. That's the beauty of standards 😉

Copy link

TFrascaroli commented Nov 22, 2020

@peterbajwa Base64 is base64, no matter who produces it or who consumes it. That's the beauty of standards 😉

Copy link

YATACOBLAS commented Apr 15, 2021

thank you !!! @jonleighton

Copy link

Github743 commented Oct 3, 2021

When I am uploading a 10 MB file some how it is not giving any conversion. Here is what I am working on

Copy link

yongky-utama commented Nov 10, 2021

I know it's been a long time, but as I tried to figure out why it didn't work, I found that in your loop, buffer[i] is always undefined.
It turned out that you just need to convert it to Uint8Array first.

function uint8ToBase64( buffer ) {
    var binary = '';
    var bytes = new Uint8Array(buffer);
    var len = bytes.byteLength;
    for (var i = 0; i < len; i++) {
        binary += String.fromCharCode(bytes[i]);
    return window.btoa( binary );

Which is the same as the top answer in this stack overflow topic: ( )

It's correct as far as I can tell, as modern browsers allows '\x00' characters in JS Strings.

Copy link

EmanH commented Feb 25, 2022

Updated benchmarks:

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment