Base64 is a group of similar binary-to-text encoding schemes that represent binary data in an ASCII string format by transforming it into a radix-64 representation. The term Base64 originates from a specific MIME content transfer encoding.

Base64 encoding schemes are commonly used to encode binary data for storage or transfer over media that can only deal with ASCII text (or some superset of ASCII that still falls short of accepting arbitrary binary data). This ensures that the data remains intact without modification during transport. Common applications of Base64 include:

• Email via MIME
• Storing complex data in XML
• Encoding binary data so that it can be included in a data: URL

Base64 characters

When the term "Base64" is used on its own to refer to a specific algorithm, it typically refers to the version of Base64 outlined in RFC 4648, section 4, which uses the following alphabet to represent the radix-64 digits, alongside = as a padding character:

ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/

URL and filename safe Base64

A common variant of this definition allows only characters that are safe to use in filenames and URL values. This version, defined in RFC 4648, section 5, omits the padding and replaces + and / with - and _.

You don't need this encoding if you are not putting the data in a path segment or query parameter — for example, data URLs have neither and can use the standard Base64 encoding.

Encoded size increase

Each Base64 digit represents 6 bits of data (64 = 2⁶). So, three 8-bit bytes of the input string/binary file (3×8 bits = 24 bits) can be represented by four 6-bit Base64 digits (4×6 = 24 bits).

This means that the Base64 version of a string or file is typically roughly a third larger than its source (the exact size increase depends on various factors, such as the absolute length of the string, its length modulo 3, and whether padding characters are used).

Last chunk

The base64 string can be partitioned into chunks of 4 characters, where the last chunk may have fewer than 4 characters. The last chunk may be padded with = characters so it's exactly 4 characters long. Excluding padding characters, the last chunk may be one of the following:

• 2 characters: encodes 12 bits representing 1 byte (8 bits) of data
• 3 characters: encodes 18 bits representing 2 bytes (16 bits) of data
• 4 characters: encodes 24 bits representing 3 bytes (24 bits) of data

In the first two cases, the characters may have 4 or 2 extra trailing bits that don't represent any data. In this case, RFC 4648 requires encoders to set these bits to zero and decoders to optionally throw an error if they are not zero. For example, if the encoded data is a single byte 0b01010101, then it needs two characters 0b010101 (V) and 0b010000 (Q), where the second character has 4 trailing bits set to zero. Decoding VR== (where the second character represents 0b010001) technically results in the same byte 0b01010101, but the decoder may throw an error due to the trailing bits not being zero.

JavaScript support

The Uint8Array class provides the Uint8Array.fromBase64(), Uint8Array.prototype.toBase64(), and Uint8Array.prototype.setFromBase64() methods for conversion to/from base64 strings.

Browsers also natively provide two JavaScript functions for decoding and encoding Base64 strings:

• Window.btoa() (also available in workers): creates a Base64-encoded ASCII string from a string of binary data ("btoa" should be read as "binary to ASCII").
• Window.atob() (also available in workers): decodes a Base64-encoded string ("atob" should be read as "ASCII to binary").

See Also

• JavaScript APIs:
  • Window.atob() (also available in workers)
  • Window.btoa() (also available in workers)
  • Uint8Array
• Data URLs
• Base64 on Wikipedia
• Base64 Algorithm described in RFC 4648