At a glance
- Eleven styles: UPPERCASE, lowercase, Sentence, Title, camelCase, snake_case, kebab-case, dot.case, CONSTANT_CASE, transliterate, URL slug.
- One switch: picking a style replaces the previous one, and tapping it again returns the raw text.
- Per-line affixes: bullets, numbering,
>quoting, trailing punctuation, or your own prefix and suffix. - Quote repair both ways: curly to straight ASCII, or straight to typographic, with guillemets for Cyrillic.
- Live counts: characters, words, lines and an approximate token count on every text item.
Eleven styles, and they toggle
The case chips behave as one exclusive group. Flip from camelCase to snake_case directly, with no reset in between, because the transform always applies to the original clipboard text rather than to the last result.
Affixes and cleanup steps work the other way. They stack, so numbering the lines and then adding a trailing comma is two taps on the same paste. The backspace chip removes the last step, and the dash resets everything.
| Style | hello world becomes |
|---|---|
| UPPERCASE | HELLO WORLD |
| lowercase | hello world |
| Sentence case | Hello world |
| Title Case | Hello World |
| camelCase | helloWorld |
| snake_case | hello_world |
| kebab-case | hello-world |
| dot.case | hello.world |
| CONSTANT_CASE | HELLO_WORLD |
| Transliterate | привет мир becomes privet mir |
| URL slug | Hello World becomes hello-world |
For code, the five identifier styles move a variable name between conventions without retyping it. Transliterate rewrites Cyrillic in Latin letters for filenames and identifiers, which is the chip I use more than I expected to.
For writing, Sentence and Title fix a headline pasted in the wrong style, and URL slug turns a post title into an address. Title Case follows simple word rules. It has never read your publication's style guide, so a house rule about short prepositions still needs a human.
Lists, quotes and encoders in the same row
Affixes apply down every line at once. Bullet turns apple and banana into - apple and - banana. Number produces 1. and 2., renumbered as you go. Quote prefixes each line with > for Markdown. Trailing punctuation adds a comma, semicolon or period to every line, which is how I build a SQL IN list.
Curly quotes break things quietly. They fail in code, in CSV imports and in search fields, and Word or Google Docs put them back through autocorrect. One tap converts "smart" 'quotes' to straight ASCII marks. The reverse direction produces typographic quotes, and guillemets for mostly-Cyrillic text.
Three encoders sit beside the case chips: URL-encode turns a b & c into a%20b%20%26%20c, Base64 turns hello into aGVsbG8=, HTML-escape turns <b> into <b>. Decoding is automatic. Copy a Base64 blob, a percent-encoded string or a JWT and Klipto offers the matching decode instead, as covered in JSON and code transforms.
The count, tokens included
Every text item carries four running numbers: characters, words, lines and an approximate token count. The token figure needs one caveat. Klipto estimates it as characters divided by four, which tracks real tokenizers closely for English prose and drifts on code, other alphabets and long runs of digits. It answers whether a prompt will fit. It will not settle an API bill.
That number in the paste bar is why I stopped keeping a counter tab open. If you want one anyway, try it free in your browser or use the token counter. Klipto itself is a native app of about 11 MB, offline, one-time $19.99 with a free tier and updates included.