About

Base58 encoding maps arbitrary byte sequences onto an alphabet of 58 visually unambiguous characters, deliberately excluding 0 (zero), O (uppercase o), I (uppercase i), and l (lowercase L) to eliminate transcription errors in manual copying. The encoding converts a byte array into a base-58 numeral by treating the input as a single big-endian unsigned integer, then performing repeated division by 58. Each leading 0x00 byte maps to the character 1 (index zero), preserving information that pure numeric conversion would discard. A single misread character in a wallet address or API key can route funds irreversibly to a dead address or fail authentication silently.

This tool implements the full BigInt-based Base58 algorithm client-side with no server round-trips. It supports both the Bitcoin alphabet (123456789ABCDEFGHJKLMNPQRSTUVWXYZ abcdefghijkmnopqrstuvwxyz) and the Flickr variant (which swaps uppercase and lowercase). Note: this encoder operates on raw UTF-8 bytes. It does not append a Base58Check checksum. For checksum-protected encoding, a SHA-256 double-hash step is required separately.

Formulas

The encoding converts a byte array B of length n into a big-endian unsigned integer, then extracts Base58 digits by modular arithmetic:

N = n−1∑i=0 B_i ⋅ 256ⁿ⁻¹⁻ⁱ

Then repeatedly divide by 58:

digit_k = N mod 58 , N ← N58

Digits are collected in reverse order. Each leading 0x00 byte in the original array prepends the character at alphabet index 0 (which is 1 in the Bitcoin alphabet).

Where: B_i = the i-th byte of the UTF-8 encoded input string. N = the integer representation of the full byte array. digit_k = the k-th Base58 character (mapped via the alphabet lookup). The output length is approximately log₅₈(256ⁿ) ≈ n × 1.3656.

Reference Data

Index	Bitcoin Char	Flickr Char	Index	Bitcoin Char	Flickr Char
0	1	1	29	W	w
1	2	2	30	X	x
2	3	3	31	Y	y
3	4	4	32	Z	z
4	5	5	33	a	A
5	6	6	34	b	B
6	7	7	35	c	C
7	8	8	36	d	D
8	9	9	37	e	E
9	A	a	38	f	F
10	B	b	39	g	G
11	C	c	40	h	H
12	D	d	41	i	J
13	E	e	42	j	K
14	F	f	43	k	L
15	G	g	44	m	M
16	H	h	45	n	N
17	J	i	46	o	P
18	K	j	47	p	Q
19	L	k	48	q	R
20	M	m	49	r	S
21	N	n	50	s	T
22	P	o	51	t	U
23	Q	p	52	u	V
24	R	q	53	v	W
25	S	r	54	w	X
26	T	s	55	x	Y
27	U	t	56	y	Z
28	V	u	57	z	0

Frequently Asked Questions

Base64 uses 64 characters including "+", "/", and "=" for padding. Base58 uses 58 characters, deliberately removing 0, O, I, l, +, and / to avoid visual ambiguity and URL-unsafe characters. Base58 output is roughly 37% longer than the raw binary input, compared to ~33% for Base64, but it is far safer for manual transcription. Base58 also has no padding characters.

When you convert a byte array to a single integer, leading 0x00 bytes are mathematically insignificant - they vanish. Bitcoin addresses and other protocols rely on exact byte-level fidelity. The algorithm counts leading 0x00 bytes before conversion and prepends one "1" character per zero byte to the final output. Without this step, decoding would yield a shorter byte array than the original.

Both alphabets contain the same 58 characters. The Bitcoin alphabet orders uppercase letters before lowercase (123456789ABCDEFGHJKLMNPQRSTUVWXYZabcdefghijkmnopqrstuvwxyz). The Flickr alphabet reverses this, placing lowercase before uppercase (123456789abcdefghijkmnopqrstuvwxyzABCDEFGHJKLMNPQRSTUVWXYZ). The numeric encoding is identical; only the character mapping differs.

No. Base58Check appends a 4-byte checksum (derived from a double SHA-256 hash of the payload with a version prefix) before Base58-encoding the result. This tool performs raw Base58 encoding without any checksum. If you need Base58Check (e.g., for Bitcoin addresses), you must compute the SHA-256 double-hash and append the first 4 bytes to your data before encoding.

This tool encodes the UTF-8 byte representation of the input text. If your input contains multibyte characters (e.g., emoji, CJK), the encoder correctly processes all UTF-8 bytes. For raw binary data (arbitrary hex), you would need to convert the hex string to a byte array first. The tool processes text as-is via the TextEncoder API.

The tool enforces a limit of 100,000 characters. For very large inputs, the BigInt arithmetic may take a few hundred milliseconds. Base58 encoding is inherently O(n²) for large inputs because each division step processes the entire number. For data exceeding a few kilobytes, consider chunked encoding or alternative formats like Base64.

Index	Bitcoin Char	Flickr Char	Index	Bitcoin Char	Flickr Char
0	1	1	29	W	w
1	2	2	30	X	x
2	3	3	31	Y	y
3	4	4	32	Z	z
4	5	5	33	a	A
5	6	6	34	b	B
6	7	7	35	c	C
7	8	8	36	d	D
8	9	9	37	e	E
9	A	a	38	f	F
10	B	b	39	g	G
11	C	c	40	h	H
12	D	d	41	i	J
13	E	e	42	j	K
14	F	f	43	k	L
15	G	g	44	m	M
16	H	h	45	n	N
17	J	i	46	o	P
18	K	j	47	p	Q
19	L	k	48	q	R
20	M	m	49	r	S
21	N	n	50	s	T
22	P	o	51	t	U
23	Q	p	52	u	V
24	R	q	53	v	W
25	S	r	54	w	X
26	T	s	55	x	Y
27	U	t	56	y	Z
28	V	u	57	z	0

Index	Bitcoin Char	Flickr Char	Index	Bitcoin Char	Flickr Char
0	1	1	29	W	w
1	2	2	30	X	x
2	3	3	31	Y	y
3	4	4	32	Z	z
4	5	5	33	a	A
5	6	6	34	b	B
6	7	7	35	c	C
7	8	8	36	d	D
8	9	9	37	e	E
9	A	a	38	f	F
10	B	b	39	g	G
11	C	c	40	h	H
12	D	d	41	i	J
13	E	e	42	j	K
14	F	f	43	k	L
15	G	g	44	m	M
16	H	h	45	n	N
17	J	i	46	o	P
18	K	j	47	p	Q
19	L	k	48	q	R
20	M	m	49	r	S
21	N	n	50	s	T
22	P	o	51	t	U
23	Q	p	52	u	V
24	R	q	53	v	W
25	S	r	54	w	X
26	T	s	55	x	Y
27	U	t	56	y	Z
28	V	u	57	z	0

Index	Bitcoin Char	Flickr Char	Index	Bitcoin Char	Flickr Char
0	1	1	29	W	w
1	2	2	30	X	x
2	3	3	31	Y	y
3	4	4	32	Z	z
4	5	5	33	a	A
5	6	6	34	b	B
6	7	7	35	c	C
7	8	8	36	d	D
8	9	9	37	e	E
9	A	a	38	f	F
10	B	b	39	g	G
11	C	c	40	h	H
12	D	d	41	i	J
13	E	e	42	j	K
14	F	f	43	k	L
15	G	g	44	m	M
16	H	h	45	n	N
17	J	i	46	o	P
18	K	j	47	p	Q
19	L	k	48	q	R
20	M	m	49	r	S
21	N	n	50	s	T
22	P	o	51	t	U
23	Q	p	52	u	V
24	R	q	53	v	W
25	S	r	54	w	X
26	T	s	55	x	Y
27	U	t	56	y	Z
28	V	u	57	z	0