Hex Converter

A hex converter earns its keep when you do not yet know what the hex in front of you is. Is 1F4 a number, the short colour #1F4, or two bytes copied backwards out of a memory dump? This one does not make you decide first. Type or paste the value once and every reading comes out together: hexadecimal, decimal, binary, octal, the byte count, both byte orders, the colour when the digits form one, the text when the bytes decode as UTF-8, and a hex dump. All of the arithmetic runs in BigInt, so a 128 bit UUID converts as exactly as a single byte.

How hexadecimal works

Hexadecimal is base 16. It counts 0 to 9 the way decimal does, then carries on with A, B, C, D, E and F for ten to fifteen, so a single hex digit holds sixteen values. Each place is worth sixteen times the place to its right.

To convert hex to decimal, multiply each digit by its place value and add the results. 1F4 is 1 × 256 + 15 × 16 + 4 × 1, which is 256 + 240 + 4, or 500. The same method gives 3B = 59, 2AA = 682 and E7A9 = 59,305. Upper and lower case letters mean the same thing, so 1f4 and 1F4 are one number.

To convert decimal to hex, divide by 16, write down the remainder, and repeat on the quotient until nothing is left. 500 ÷ 16 is 31 remainder 4, 31 ÷ 16 is 1 remainder 15, which is F, and 1 ÷ 16 is 0 remainder 1. Read the remainders from the bottom up and you have 1F4. The working panel under the ladder prints the right method for whatever you enter, digit by digit.

Why computers write numbers in hex

Sixteen is two to the fourth power, so one hex digit stands for exactly four bits, a unit called a nibble, and two hex digits make exactly one byte. The fit is perfect: FF is the largest value a byte holds, 255, and the number after it, 100, is 256. Binary spells that same byte as 11111111, eight characters that are easy to miscount, and decimal needs one to three digits that never line up with byte boundaries. Hex is the shortest notation that still shows where every byte starts and ends, which is why memory addresses, hashes, colour codes and network addresses are all written in it. When you need to see or flip individual bits, or work out how a negative number is stored, the binary converter spells out every one.

Read the prefix before the digits

Hex digits look like decimal digits, so written hex nearly always carries a marker. 0x is the C convention, kept by JavaScript, Python and most languages that borrowed C's syntax. Intel assembly puts an h after the number and adds a leading 0 when the first digit is a letter, as in 0FFh. 6502 and Motorola assembly, Pascal and Delphi put a $ in front. Ada and VHDL wrap the digits as 16#5A3#. A # in front means a colour in CSS or HTML. U+ marks a Unicode code point, &#x and a closing semicolon make an HTML character reference, \x marks one byte inside a string literal, and % marks one byte in a URL. The converter strips all of these, along with spaces, colons and hyphens between bytes, so you can paste a value exactly as you found it. The reference table above sorts the prefixes by what they tell you: a number, a colour, text or an address.

Hex colour codes

A CSS hex colour is three bytes, red, green and blue, each from 00 to FF. #FF6C00 is rgb(255 108 0): full red, a little under half green and no blue, a strong orange. Six digits give 16,777,216 possible colours. Two other lengths are easy to misread. Three digits double each one, so #F09 is #FF0099 and #123 is #112233, not #010203. Four and eight digits add an alpha channel on the end, where 00 is fully transparent and FF fully opaque. In the middle of the 1990s, when many screens showed only 256 colours, designers kept to a web safe palette built from the steps 00, 33, 66, 99, CC and FF, 216 colours in all.

Whenever the hex you type has 3, 4, 6 or 8 digits the converter adds a colour row with a swatch and the rgb() values. That is why the default value, 1F4, shows up as a vivid green: read as a colour, #1F4 is #11FF44. To pick a colour or look one up by name rather than decode one, use the RGB Color Codes tool.

Text to hex, and hex back to text

Text only becomes hex through an encoding, and on the web that encoding is UTF-8: W3Techs counted it on 99.1% of the websites whose encoding it could identify in September 2026. UTF-8 stores each of the 128 ASCII characters in one byte with the same value ASCII gives it, so H is 48, e is 65 and Hello is 48 65 6C 6C 6F. Every other character takes two, three or four bytes. é is code point U+00E9 and is stored as C3 A9; £ is C2 A3; € is E2 82 AC. So a word with an accent has more bytes than letters, and a program that wrongly reads one byte per character turns café into café.

Switch Read it as to Text and the ladder shows the UTF-8 bytes, the Unicode code points, the same bytes as \x and % escapes, and a hex dump. In hex mode the converter tries the reverse on every value and adds an As text row only when the bytes form valid, printable UTF-8, so an ordinary number never turns into a line of junk characters. For the full chart of the 128 ASCII codes, see the ASCII table.

Byte order: why memory dumps look backwards

A number wider than one byte has to be laid out in memory one byte at a time, and machines disagree about which end goes first. Big endian puts the most significant byte first, the way people write numbers. Little endian puts the least significant byte first. The 32 bit value 0x0A0B0C0D sits in memory as 0A 0B 0C 0D on a big endian machine and as 0D 0C 0B 0A on a little endian one. x86, x86-64 and most ARM systems are little endian, while Internet protocols send the big end first, which is why big endian is also called network byte order. Danny Cohen introduced the names in 1980, taking them from the two factions in Gulliver's Travels that argue over which end of an egg to crack.

In practice this means that when a debugger shows F4 01 00 00 at an address, the number stored there is 0x000001F4, which is 500. Set Byte order typed to little endian, paste the bytes as they appear, and the converter reverses them before it converts. The ladder always prints both orders, and Pad to width fills a value out to 8, 16, 32, 64 or 128 bits so you can see it laid out the way a register or a struct field would hold it.

Long values stay exact

A converter built on ordinary JavaScript numbers holds whole numbers exactly only up to 9,007,199,254,740,991, 2 to the 53rd power minus 1, written 1FFFFFFFFFFFFF in hex. Above that, some integers simply cannot be stored and the tool quietly hands back a neighbour: 9,007,199,254,740,993 comes back as 9,007,199,254,740,992. Other tools refuse long input outright; one popular pair of converters stops at 16 hex characters and 19 decimal digits, the 64 bit ceiling. This page uses BigInt, whole numbers with no fixed size, so a 32 digit UUID, a 64 character SHA-256 hash or a 40 character Ethereum address converts digit for digit.

Identifiers written in hex

Several everyday identifiers are simply long numbers written in hex with fixed grouping. A UUID is 128 bits, written as 32 hex digits in groups of 8, 4, 4, 4 and 12, 36 characters with the hyphens; RFC 9562 replaced RFC 4122 as its specification in May 2024. A MAC address is 48 bits, six pairs of hex digits separated by colons or hyphens, and the first three pairs identify the manufacturer. An IPv6 address is 128 bits in eight groups of four hex digits, and its standard text form uses lowercase letters, drops leading zeros and shortens the longest run of zero groups to a double colon. Paste a UUID or a MAC address straight in and the separators are ignored, so you see the identifier as one number and as its bytes. For IPv6, write every group out as four digits first, because the shortened form leaves out zeros the converter cannot guess.

Hex dumps and file signatures

A hex dump shows raw bytes as text: an offset on the left, usually 16 bytes per line in hex, and on the right the same bytes as characters, with anything outside the printable range 20 to 7E shown as a dot. The Unix tools xxd, hexdump and od all print this layout, and the converter adds a dump to the ladder for any value of four bytes or more. Dumps are how you recognise a file from its first few bytes, its signature or magic number. A PNG image begins 89 50 4E 47 0D 0A 1A 0A, with the letters PNG readable in the dump. A PDF begins 25 50 44 46 2D, which reads %PDF-. A ZIP archive begins 50 4B 03 04, PK. ELF programs on Linux begin 7F 45 4C 46, and Java class files begin CA FE BA BE.

What this converter leaves to other tools

It converts whole numbers and bytes. Hex fractions exist, 0.8 is one half because 8 is half of 16, but they are rare outside the classroom. Floating point values are stored in their own IEEE 754 bit layout, so the hex of a float is not the hex of the whole number it looks like; a float converter is the right tool there. Adding and subtracting in hex, where a borrowed 1 is worth 16 rather than 10, is a job for a hex calculator. Negative numbers keep their minus sign here, and how a machine stores them in a fixed width, two's complement, is covered on the binary converter.