Text to binary, formatted the way you need it

Type it into the box at the top and read the digits underneath — Ana comes out as 01000001 01101110 01100001 on the default settings. Each group of eight is one letter, and capitals carry different numbers from lowercase, so Ana and ana do not produce the same digits. For a bracelet, a tattoo or anywhere the spacing would look wrong, set the separator to nothing between groups and it becomes one unbroken string.

Because UTF-8 spends four bytes on every character numbered above 65,535, and all emoji are numbered up there. Unicode is far bigger than the 256 numbers a single byte can hold, so UTF-8 stretches: one byte for the original ASCII characters, two for accented Latin, three for most Chinese, Japanese and Korean, four for emoji and the rarer scripts. A family emoji is stranger still, because it is three separate people welded together by two invisible zero-width joiners — five characters and 18 groups for what looks like one picture, which is why the total above the box jumps further than you expect.

The 7-bit setting drops the leading zero that every ASCII character carries inside an eight-digit group, so A becomes 1000001 instead of 01000001. Seven is the honest width for ASCII, whose numbers never pass 127, and it makes the whole output an eighth shorter. It also refuses anything outside ASCII: put an accented letter in and the converter stops and names the character and its position, because no 7-bit number exists for it.

Yes — choose nothing between groups and the output becomes one continuous string. Keep the width at eight bits if you do, because an unbroken string can only be read back when every character occupies the same number of digits. With the trimmed width the groups vary in length, so a run like 100100011001 has no single correct reading: it is a good party trick and a hopeless way to send a message.

Yes, a space is character number 32 and comes out as 00100000. Every punctuation mark has a number too, and a line break is number 10, which the table below shows as line break rather than as an empty cell. This is why the phrase Hi there produces 8 groups rather than 7, and why counting groups is a faster way to count characters than counting them by eye.

Exactly what the output box holds: the digits, with the separator and the line breaks you chose, saved as a plain text file called text-to-binary.txt. It is assembled in the browser from the text already on your screen and handed straight to your downloads folder, so nothing travels to a server to make it. Any text editor opens it, phones included, although a single very long line wraps awkwardly on a small screen — pick 8 groups per line before saving if that matters.

Almost always because of padding, separators or encoding, which are the three things the menus here put in your hands. A site that pads to eight bits and one that stops at seven produce visibly different output for identical text, and a site built on an old single-byte table gives one group for an accented letter where UTF-8 gives two. Match a specific target by setting the width and separator to whatever it uses; to find out what a mystery string actually says, decode it on the binary translator instead.