Base64 decoding step by step

Decoding reverses the encoding steps. Let’s decode Q2F0 and Q2F0cw==.

Example 1: “Q2F0”

Step 1 — characters to values using the Base64 alphabet:

Step 2 — join the bits and split into bytes (8 bits):

010000110110000101110100
01000011 | 01100001 | 01110100
   67    |    97    |   116

Step 3 — bytes to characters with the ASCII table: 67 → C, 97 → a, 116 → t.

Result: Q2F0 → Cat

Example 2: “Q2F0cw==” (with padding)

The first block decodes to Cat as above. The second block is cw==:

c       w
011100  110000

Two = mean this block holds only one byte. Take the first 8 of the 12 bits and discard the rest (which must be zeros in strict decoding):

01110011 | (0000)
   115 → "s"

Result: Q2F0cw== → Cats

What can go wrong

• Invalid characters (spaces, -, _, %) have no value in the standard alphabet. Strict decoders stop; lenient ones skip them. Our validator shows their exact position.
• Wrong length: a valid unpadded length is never 4n + 1, because one leftover character holds only 6 bits — not enough for a byte.
• Non-zero leftover bits, like Q2F0cx==, are technically invalid but most decoders accept them and decode the same bytes.
• Text looks garbled after decoding: the bytes are correct but displayed with the wrong character set. Use character encoding detection.

Try your own value with the Base64 decoder.