Why does a Python dunder name like __init__ come out as just "init" instead of keeping the double underscores?

Leading, trailing, and doubled separators collapse to nothing during splitting, the parser is only trying to find word boundaries, not preserve every separator character verbatim. __init__ produces the single word "init", the same result you'd get from converting plain "init", so the double-underscore convention itself doesn't survive a round trip through this tool.

Does a number stick to the word next to it, or become its own separate word?

It sticks to whichever letters it's directly adjacent to rather than splitting off on its own: version2Point0 becomes the two words "version2" and "point0", not four separate tokens. Only a lowercase-to-uppercase letter transition creates a new split point; digits don't trigger one by themselves.

What does "auto-detect" actually mean here, does the tool guess which case I want to convert to?

No, there's nothing to guess on the output side, all nine cases are always generated together regardless of what you paste. "Auto-detect" refers only to correctly splitting whatever mixed format you typed, camelCase, snake_case, spaces, or any combination, into individual words before converting, not to picking one target format for you.

Text Case Converter: camelCase, snake

Inputpaste or type · or drop a .txt file

Browse 3 words

camelCasehelloWorldExample

PascalCaseHelloWorldExample

snake_casehello_world_example

SCREAMING_SNAKE_CASEHELLO_WORLD_EXAMPLE

kebab-casehello-world-example

Title CaseHello World Example

UPPERCASEHELLO WORLD EXAMPLE

lowercasehello world example

dot.casehello.world.example

Auto-detects camelCase, PascalCase, snake_case, kebab-case, spaces and mixed input

When to use each case

Seven conventions cover almost everything you'll touch as a developer.

camelCaseJS/TS variables, functions, object keys

PascalCaseClass names, React components, TS types

snake_casePython, Ruby, database columns

SCREAMING_SNAKEConstants, environment variables

kebab-caseCSS classes, HTML attributes, URL slugs

Title CaseHeadings and proper nouns

dot.caseConfig keys (e.g. logging.level)

How auto-detection works

The converter parses any input format into a sequence of words, then re-formats them into every target case. It splits on spaces, underscores, hyphens, dots, slashes, and colons.

For camelCase and PascalCase input, it inserts a split point before each uppercase letter following a lowercase one (e.g. "myVariableName" → ["my", "variable", "name"]) and handles acronyms (e.g. "parseHTMLString" splits into "parse", "html", "string" — each case then re-applies its own capitalization). Mixed inputs like "my-Variable_name" are fully supported.

Case conventions by language

The same case can carry different meaning depending on the ecosystem.

JS / TScamelCase vars/fns, PascalCase classes, SCREAMING constants

Pythonsnake_case vars/fns, PascalCase classes, SCREAMING constants

CSSkebab-case classes and custom properties (--my-color)

SQLUPPER_SNAKE keywords by convention, snake_case tables/columns

GoPascalCase exported, camelCase unexported

Rustsnake_case fns/vars, PascalCase types/traits

Frequently asked questions

Every word is lowercased during splitting, then only its first letter is capitalized when rebuilding a case that needs it, so parseHTMLString correctly splits before and after the acronym but rejoins as ParseHtmlString, not ParseHTMLString. The split point is right, the original all-caps styling of the acronym itself isn't carried through to camelCase, PascalCase, or Title Case output.