Case Converter Tool

Switch text between uppercase, lowercase, title case, sentence case, camelCase, PascalCase, snake_case, and kebab-case in seconds. Retyping a headline to fix its capitals, or renaming forty fields from one convention to another by hand, is slow work that is easy to get subtly wrong — and the mistakes are the kind nobody spots until the page is published or the build fails.

A case converter removes the typing, but the reason to use this one is the part most converters get wrong. Capitalization is not a character-by-character operation. It depends on where a word sits, what kind of word it is, which style guide you follow, whether the letters are Latin at all, and whether the word is a brand name that must never be touched. Every one of those questions has a control on this page, and the defaults answer them the way a careful editor would.

What it saves you

• Consistent formatting — one style across a document, a spreadsheet column or a whole folder of files.
• Naming conventions — move a list of identifiers between snake_case, camelCase and SCREAMING_SNAKE_CASE without touching the words themselves.
• Publication-ready headlines — title case to the exact rules of AP, APA, Chicago, MLA, the New York Times or Wikipedia.
• Fixing shouted text — an ALL CAPS paragraph becomes readable prose, with acronyms and abbreviations left alone.
• Live counts — characters, words, lines and sentences update as you type, on the input and on the result.

Seven styles are for prose and seven are naming conventions for code. The difference matters: the prose styles read a custom dictionary and can preserve terms like iPhone, while the code styles define the case of every character by definition and cannot have exceptions without ceasing to be the convention they name.

Prose styles

• UPPERCASE — every letter capital. HELLO WORLD.
• lowercase — every letter small. hello world.
• Sentence case — the first word of each sentence, like ordinary writing. Hello world.
• Title Case — major words capitalized, minor words not, following a published style guide. The Sound of Music.
• Capitalized Case — every word capitalized with no exceptions. The Sound Of Music.
• aLtErNaTiNg cAsE — alternating capitals, starting lowercase. hElLo WoRlD.
• InVeRsE cAsE — every letter's case flipped. hELLO wORLD.

Code naming conventions

• camelCase — first word lowercase, the rest capitalized, joined. helloWorld. Common in JavaScript, Java and Swift.
• PascalCase — every word capitalized, joined. HelloWorld. Class and type names in most languages.
• snake_case — lowercase words joined by underscores. hello_world. Python, Ruby and most SQL schemas.
• SCREAMING_SNAKE_CASE — uppercase words joined by underscores. HELLO_WORLD. Constants and environment variables.
• kebab-case — lowercase words joined by hyphens. hello-world. URLs, CSS class names and file names.
• Train-Case — capitalized words joined by hyphens. Hello-World. HTTP header names.
• dot.case — lowercase words joined by dots. hello.world. Configuration keys and namespaces.

There is no universal rule for which words in a title are capitalized. Every publication follows a style guide, and the guides genuinely disagree — often about the same word in the same sentence. Choosing the wrong one is not a matter of taste; a manuscript submitted to a journal that requires APA and formatted to Chicago will come back.

This tool implements six, and the differences below are exactly what separates them. Pick your guide from the rail and the result changes accordingly.

Where the guides disagree

• Preposition length — AP and APA lowercase prepositions of three letters or fewer. Chicago (18th edition) and Wikipedia lowercase them up to four. MLA lowercases every preposition, whatever its length.
• The New York Times uses a list, not a length — at, by, in, of and to are lowercase, but off, out, up, no, nor, not and so are capitalized — which no other guide here does.
• Coordinating conjunctions — five of the six lowercase all seven of for, and, nor, but, or, yet and so. Chicago lowercases only five of them: yet and so are capitalized.
• The last word — every guide here capitalizes it except APA, which has no such rule — so an APA title really does end "...Service You Can Count on".
• Infinitive "to" — AP capitalizes it in an infinitive ("How To Write a Book") and lowercases it as a preposition ("Back to Work"). The other five lowercase it always.
• Hyphenated compounds — AP, APA and Wikipedia capitalize both parts. Chicago and MLA apply the minor-word rules to the second part, giving "Follow-up" where the others give "Follow-Up". The New York Times lowercases the second part only after a two- or three-letter prefix across doubled vowels: "Co-operation", but "Co-Author".
• Latin phrases — Chicago capitalizes the preposition inside a Latin expression used adverbially, so "In Situ" rather than "in Situ".

One honest limit applies to every tool in this category, including the ones that charge for it: whether "off" is a preposition or an adverb depends on the sentence, and no word list can tell. Each word is given its usual role in a title, which is right for the overwhelming majority of real headlines. When it is wrong for yours, put the word in the custom dictionary with the spelling you want and it wins over every rule.

Four steps, and only the first two are needed for most conversions.

1. Add your text

   Paste or type into the input box, or drop a .txt, .md, .csv, .tsv or .log file onto the page. Anything longer than about 128,000 characters is queued as a job instead of filling the input box, so the page stays responsive.

2. Choose a case style

   Pick one of fourteen styles. Seven are for prose — UPPERCASE, lowercase, Sentence case, Title Case, Capitalized Case, aLtErNaTiNg and InVeRsE — and seven are naming conventions for code: camelCase, PascalCase, snake_case, SCREAMING_SNAKE_CASE, kebab-case, Train-Case and dot.case.

3. Set the rules that apply

   For Title Case, choose the style guide your publication uses. Add any terms whose exact spelling must survive — iPhone, NASA, macOS — to the custom dictionary, and set the casing language if your text is Turkish, Azerbaijani, Lithuanian, Greek or German.

4. Copy or download

   Short text converts as you type. Files and long pasted text convert when you press Convert, and a batch downloads as a single ZIP. Everything runs on your device; no text is uploaded.

Sentence case looks like the simplest style and is the one most tools implement incorrectly. The naive version lowercases the whole input and then capitalizes the letter after every full stop, which quietly does four wrong things at once.

What a correct sentence case has to know

• The pronoun I — lowercasing everything first turns every I in the document into i. It stays capital here, along with I'm, I've, I'll and I'd.
• Abbreviations — Dr., e.g., i.e., Inc., Fig., Vol. and around eighty more end in a full stop without ending a sentence. A new sentence does not start after any of them.
• Initials — a single capital letter followed by a full stop is an initial. J. R. R. Tolkien keeps its initials and does not restart the sentence three times.
• Decimals and domains — version 3.5 and example.com contain full stops that are not sentence endings, and neither restarts the capitalization.
• Sentences that start with a number — in "3 apples are enough", the first word is the number, so "apples" is not capitalized. Tools that hunt for the first letter get this backwards.

Acronyms and brand names survive too, as long as 'Preserve capitalized terms' is on: any word carrying a capital after its first letter is left exactly as written. The one deliberate exception is a line entirely in capitals, where preserving them would make the tool do nothing to the input people most often bring to it.

Moving a set of identifiers from one convention to another is the most common developer use of a case converter, and the accuracy of the result depends entirely on where the tool decides one word ends and the next begins.

How word boundaries are found

• Separators — spaces, underscores, hyphens, dots and slashes all end a word. user_profile-image is three words.
• Camel humps — a capital after a lowercase letter starts a new word. fooBar is foo and Bar.
• Acronym runs — a capital that is followed by a lowercase letter ends a run of capitals. HTTPServer is HTTP and Server, and parseJSONData is parse, JSON and Data. This is the rule most converters omit, and without it those become httpserver and parsejsondata.
• Digits stay attached — utf8 is one word, but utf8Encode is two — the capital is the boundary, not the digit.
• Apostrophes stay inside words — don't is one word and becomes dont in an identifier, not don and t. A closing quote in dogs' toys is still a separator.

Two options worth knowing

• Convert line by line — each line becomes its own identifier and the line breaks survive. Use it for a column of field names or a list of variables; leave it off to collapse a whole block into one identifier.
• Keep existing separators — the original separators are kept and only the words change case, which is what you want for a path or a structured key. src/my_module/file.txt becomes Src/My_Module/File.Txt in PascalCase rather than SrcMyModuleFileTxt.

Most case converters are written against the twenty-six letters of the English alphabet, and the failure is silent rather than loud: they do not report an error, they return a shorter word or an empty result.

What is handled here

• Accented Latin — Café stays Café. A tool that looks for A-Z drops the é and returns Caf, and nothing about that looks like a failure.
• Other alphabets — Cyrillic, Greek, Hebrew, Arabic, Devanagari and CJK text converts correctly. In an A-Z tool, camelCase of Cyrillic text returns an empty string.
• Turkish and Azerbaijani — i uppercases to the dotted İ and I lowercases to the dotless ı. Set the casing language and the mapping follows the language rather than English.
• German ß — uppercases to SS, which is two letters from one — the reason nothing in this tool assumes a case change preserves the length of a word.
• Combining accents — an e followed by a combining acute is one word, not a word plus a stray mark, so the accent cannot migrate onto the previous word.
• Emoji and rare scripts — characters outside the basic plane are handled as whole characters. An emoji counts as one character, not two, and is never split in half.

The input box is for text you paste or type. For anything larger, drop the file instead — .txt, .md, .markdown, .csv, .tsv and .log are accepted, and UTF-8, UTF-8 with a byte-order mark and UTF-16 with a mark are all detected and written back in the encoding they arrived in.

What happens with a large document

• A big paste becomes a job — text longer than about 128,000 characters is queued rather than typed into the input box, because a form control holding megabytes of text is what makes a page freeze — not the conversion itself.
• Long conversions run in the background — the page stays responsive and Cancel works immediately, however large the document.
• Line endings survive — a file that uses CRLF comes back using CRLF. Nothing is silently normalised.
• Formats that are refused — JSON, XML and HTML are not accepted, because case carries structure in all three and lowercasing a key or a tag name breaks the file. Word, PDF and spreadsheet files are refused with a link to the right converter.
• Batches download as one archive — convert up to 50 files in a run and take them as a single ZIP, with duplicate file names kept distinct rather than overwriting one another.

For the formats this tool does not open, these convert them to plain text first: Word to Text, PDF to Text, Excel to CSV, Image to Text.

A few things worth knowing before converting anything you cannot easily undo.

Getting a clean result

• Convert a short sample first when the text matters — a headline, or the first ten rows of a column — and check the two or three words you already suspect.
• Add brand names, product names and acronyms to the custom dictionary before a long run, not after it.
• Pick the style guide your destination requires rather than the one that looks best; the differences are small and the requirement is not negotiable.
• For identifiers, check what happened to the acronyms — that is where naming conversions go wrong.
• Keep the original. The tool never modifies your file; it produces a new one.

Where a case converter is the wrong tool

• Trademarks and legal text whose capitalization is fixed by contract. Use the dictionary, or do not convert at all.
• Source code, as opposed to a list of identifiers: changing case across a whole file will change keywords and string literals too.
• Rich text. Paste as plain text; bold, links and colour are not carried through.
• Scripts with no upper and lower forms — Chinese, Japanese, Korean, Arabic, Hebrew and Thai are unchanged by the prose styles, which is correct rather than a failure.

No text is uploaded. The conversion happens on your own machine, which is why there is no file size to negotiate with a server and no queue to wait in — and why pasting an unreleased product name, a client's draft or a column of customer records is safe here in a way it is not on a tool that posts your text to an API. Free converts one item at a time, up to 10 MB. Pro lifts that to whatever your device can handle and up to 50 files in a single run. Both produce identical bytes: case conversion is lossless, so there is no quality tier to sell.

• Text inverter
• Word counter
• Text analyzer
• Text merger
• Text trimmer
• Remove duplicate lines
• Sort text lines
• Split text
• Text statistics
• All text tools