What this tool does
It turns a title or phrase into a URL slug: the readable last part of a web address, such as my-first-post in /blog/my-first-post. Enter one title per line and every line gets its own slug. The tool lowercases the text, removes accents and punctuation, joins the words with the separator you choose, and removes separators from the start and end. Runs of punctuation and spaces become a single separator, so you never get -- in the middle.
The rest of this page has two parts: what the tool does (functionality), and general guidance about slugs (advice, not something the tool enforces).
How it works
| Input | Option | Slug |
|---|---|---|
10 Tips for Better Sleep! | Default | 10-tips-for-better-sleep |
Don't Panic: A Developer's Guide | Default (apostrophes removed) | dont-panic-a-developers-guide |
Café crème brûlée | Default (accents removed) | cafe-creme-brulee |
Müller & Söhne | German umlauts as ae/oe/ue | mueller-and-soehne |
Привет, мир! | Transliterate | privet-mir |
اردو زبان | Keep non-Latin letters | اردو-زبان |
日本語のテスト | Keep non-Latin letters | 日本語のテスト |
myFirstBlogPost | Split camelCase | my-first-blog-post |
The Art of the Deal | Remove stop words | art-deal |
Hello World | Underscore separator | hello_world |
- Apostrophes are removed rather than turned into separators, so
Developer'sbecomesdevelopers, notdeveloper-s. - Accents on Latin letters are removed (
ébecomese). Letters such asß æ ø łhave their own replacements (ss ae o l). Tick the German option forä ö üasae oe ue. - Compatibility forms are normalized: full-width letters, ligatures such as
fiand superscript digits become plain letters and digits. - Digits written in Arabic-Indic, Persian, Devanagari, Bengali or full-width forms are converted to 0-9.
- Length. The limit counts characters, not bytes. With Cut at a whole word the slug ends at the last word that fits; otherwise it is cut at the limit and the separator at the end is removed.
- Duplicates inside your list are reported. They are only detected within the lines you enter, because the tool cannot see the slugs already on your site.
Non-Latin text, Urdu and other languages
The Non-Latin text menu decides what happens to letters outside the Latin alphabet:
- Drop it. Only A-Z and 0-9 are kept. A title written entirely in Urdu gives an empty slug, and the tool says so instead of hiding it.
- Keep non-Latin letters. Letters and digits of any script are kept (Urdu, Arabic, Hebrew, Cyrillic, Greek, Devanagari, Chinese, Japanese, Korean and others) and punctuation becomes the separator. The text is normalized (NFKC) so that the same word typed in different ways gives the same slug. Latin accents are still removed in mixed text.
- Transliterate Cyrillic and Greek. These two scripts are converted letter by letter into Latin. It is a simple, general scheme (for example Cyrillic
щbecomesshchand Greekευbecomesey), not a language-specific standard such as BGN/PCGN, ISO 9 or ELOT 743, so check names against the spelling you want.
What it does not do. It does not romanize Urdu, Arabic, Hindi, Chinese, Japanese or Korean. Romanization of Urdu and Arabic needs vowels that are usually not written, and Chinese and Japanese need dictionaries. A reliable result would need a language-specific tool, so the tool keeps the original letters or drops them and leaves the choice to you.
Urdu and Arabic details. The tool removes zero-width joiners and non-joiners (used inside Persian and Urdu words), the stretching character (tatweel) and, by default, short-vowel marks such as zer, zabar and pesh, so that the same word does not produce several different slugs. An optional setting converts Arabic yeh and kaf (ي ك) to the Urdu and Persian forms (ی ک). Turn it on only when that matches your content; it would change genuine Arabic text.
Unicode URLs: what happens on the wire
A path can contain non-ASCII characters, but the address that is sent over the network is percent-encoded: each character is converted to its UTF-8 bytes and every byte is written as %XX. A browser often shows the readable form in the address bar, while logs, emails, copied links and many tools show the encoded one.
Example: the slug اردو-زبان is 9 characters and 17 bytes. Sent in a URL it becomes %D8%A7%D8%B1%D8%AF%D9%88-%D8%B2%D8%A8%D8%A7%D9%86, which is 49 characters. The table under the tool shows this for each slug.
This is separate from internationalized domain names. A domain name uses a different encoding (Punycode); the path does not. To see how a full address is split into path and query, use the URL Parser, and to encode or decode percent sequences use the URL Encoder & Decoder. Other escaping rules (JSON, HTML, shell) are in the String Escape & Unescape tool.
What makes a good slug (general guidance)
This section is advice, not a setting of the tool, and nothing here guarantees a better search position. Search engines use many signals; a slug mainly helps people and systems read and reuse a link.
- Readable and descriptive. A person should be able to guess the page from the address:
/blog/slug-generator-guiderather than/p?id=84721. - Short enough to scan. Remove filler where it does not hurt meaning, but keep words that carry it. There is no magic length.
- Lowercase and consistent. The path part of a URL is case-sensitive on many servers, so
/Aboutand/aboutcan be different addresses. Pick one style and use it everywhere. - Hyphens between words. Google's URL structure guidance recommends hyphens over underscores to separate words. Underscore and dot options are here for file names, identifiers and systems that need them.
- Stable. A slug is a public address. If you change it after publishing, add a permanent redirect from the old address.
- Free of private data. Do not put email addresses, full names of private people or internal codes in a slug. URLs end up in logs, referrers and shared links.
Readable or keyword-stuffed
| Slug | Verdict |
|---|---|
how-to-make-a-url-slug | Reads naturally and says what the page is about. |
url-slug-guide | Short and clear. |
slug-generator-url-slug-seo-slug-text-to-slug-best-slug | Stuffed: repeats the same idea, is hard to read and looks spammy. |
post-1234 | Stable but says nothing about the content. |
the-art-of-the-deal | Fine. Removing stop words gives art-deal, which is shorter but harder to read, so the tool leaves that option off by default. |
A slug should describe the page in the same words a person would use. Adding more keywords does not make it better.
For developers
- Store the slug, do not recompute it. Generate it once when the content is created, save it, and keep it even if the title changes.
- Make it unique in your database. Use a unique index and add a numeric suffix on conflict. The Number duplicates option shows the pattern, but only within the list you enter.
- Treat the slug as one path segment. The separator options never produce
/,?or#, so the result cannot split into several segments. Custom separators are limited to- _ . ~for the same reason. - Multilingual sites. Common approaches are a language prefix (
/ur/...) with a localized slug, a Latin slug shared by all languages, or an ID plus a decorative slug. Pick one and apply it consistently, and keep the language prefix as a separate path segment from the slug. - File names. A slug also works as a file name, but these rules do not check operating-system reserved names or length limits.
- Clean the list first. If your titles contain repeats, the Sort & Deduplicate Lines tool can remove them, and the Case Converter can restyle other identifiers such as variable names.
Common mistakes
- Changing slugs on a live site without redirects.
- Mixing upper and lower case, or hyphens and underscores, across pages.
- Leaving a trailing separator or double separators from copied punctuation (the tool removes these).
- Running every title through a Latin-only slug generator: a Urdu or Arabic title silently becomes an empty slug or a meaningless fragment. Choose how non-Latin text should be treated before you publish.
- Typing the same word with different Unicode forms (composed and decomposed accents, Arabic or Urdu yeh). The tool normalizes these, but slugs created elsewhere may differ.
- Putting titles with an ampersand or entities into HTML without escaping them. A slug is safe by design, but the title next to it may need the HTML Entity Encoder & Decoder.
- Assuming a generated slug will rank. It will not make a weak page rank.
How to use it
- Paste titles, one per line, or use Insert example.
- Choose the separator, case and how non-Latin text is handled. Use the extra options for stop words, camelCase and duplicates.
- Check the notes and the table, especially for empty or duplicate slugs and for percent-encoded lengths.
- Copy the output. Each line of output matches the same line of input, including blank lines.
Frequently asked questions
What is a URL slug?
The human-readable part of a URL that identifies a page, usually at the end of the path, such as my-first-post.
Should I use hyphens or underscores?
Hyphens are the usual choice for words in URLs, and Google's guidance recommends them. Use underscores only when a system requires them.
Can a slug contain Urdu or other non-Latin letters?
Yes. Choose Keep non-Latin letters. The slug stays readable in the browser, but it is longer once percent-encoded, which matters when links are copied into plain text.
Does the tool check my existing slugs?
No. It only knows the lines you enter, so check conflicts against your own site or database.
Does a slug affect search rankings?
A clear slug helps people understand a link. This tool does not claim any effect on ranking.
Privacy and local processing
This tool runs entirely in your browser. Your titles are converted by JavaScript on this page and are not sent to a server. Nothing is stored.