Free Korean Romanization Converter (로마자 변환)
Convert Hangul to Latin letters by the Revised Romanization of Korean, with the sound changes between syllables applied (국어 → gugeo, 종로 → Jongno). Names in passport style with common surname spellings, and addresses (종로구 → Jongno-gu).
Type Korean and get it in Latin letters by the Revised Romanization of Korean, with the sound changes between syllables applied: 종로 → Jongno, 신라 → Silla.
How the rules work
- Words are written as they sound: 국어 → gugeo, 백마 → baengma, 신라 → Silla, 같이 → gachi.
- Tense sounds are not marked: 압구정 → Apgujeong. ㅢ is always ui: 의정부 → Uijeongbu.
- Names: surname first, no sound changes inside the given name: 한복남 → Han Boknam.
- Addresses: a hyphen before the unit and no sound change across it: 삼죽면 → Samjuk-myeon.
🔒 Converted in this browser. Nothing you type is sent anywhere.
How It Works
Each Hangul syllable is split into its initial consonant, vowel and final consonant. The converter then applies the sound changes between neighbouring syllables (liaison, nasalisation, ㄹ assimilation, palatalisation, ㅎ aspiration) and writes the result with the Revised Romanization table.
- 1Choose what you are convertingPick "Text", "Name" or "Address" at the top.
- 2Type or paste the KoreanThe romanization appears as you type. Anything that is not Hangul is kept as it is.
- 3Pick the spellingFor a name, choose the surname spelling and the format, such as HONG GILDONG for a passport form.
- 4Copy the resultPress "Copy" next to the version you want.
Everything is converted in your browser. Nothing you type is sent anywhere.
Guide
Worked example
Open the "Name" tab and press "Try a sample", which fills in 이서연. Under "Spelling of the surname 이" you get four buttons, Lee, Yi, Rhee and I, with Lee selected. Four rows appear: "Surname and given name" Lee Seoyeon, "With a hyphen in the given name" Lee Seo-yeon, "Passport style" LEE SEOYEON and "Given name first" Seoyeon Lee. Press Rhee and every row switches to that spelling, for example Rhee Seoyeon. On the "Address" tab, the sample 서울특별시 종로구 세종대로 175 becomes Seoul-teukbyeolsi Jongno-gu Sejong-daero 175. On the "Text" tab, the sample line 신라, 종로, 같이, 좋아요 becomes silla, jongno, gachi, joayo, because the sound changes between syllables are applied.
Which settings to use, and the limits
- Pick the tab that matches what you are converting. "Text" romanizes any Korean text and keeps line breaks, punctuation, numbers and Latin letters as they are. "Name" splits a person's name into surname and given name and shows four forms. "Address" puts a hyphen before units such as 구, 동, 로 and 길. All three work as you type, and the page shows the result under the box.
- Use the two options on the Text tab for style. "Capitalise each word" starts every Korean word with a capital letter, so 안녕하세요 becomes Annyeonghaseyo. "Hyphen where it could be misread (중앙 → jung-ang)" adds a hyphen only where two syllables could be split another way, so 해운대 becomes hae-undae. The Address tab always capitalizes and does not use the hyphen option.
- For a passport or card, copy the spelling you already have. Choose the surname spelling button first, because 이 is written Lee, Yi, Rhee or I and the tool cannot know which one you use. The page warns: "If you already have a passport, bank card or visa, use exactly the spelling on it." The other forms are a convenience: the hyphen form is Gil-dong and the passport style is upper case, with the surname first.
- Names are not converted like running text. In the Name tab the given name is written syllable by syllable with no sound changes, so 한복남 gives Han Boknam. The same letters on the Text tab give hanbongnam, because the sound change between 복 and 남 is applied there. Type the surname first, as written in Korean, with at most one space after it.
- Add a space when a surname is split the wrong way. The tool reads seven two-syllable surnames as one: 남궁, 황보, 제갈, 선우, 독고, 사공 and 서문. So 황보라 gives Hwangbo and the given name Ra. If the surname is 황 and the given name is 보라, type 황 보라 with a space and you get Hwang Bora. The built-in list has the usual spellings for 95 surnames, and for any other surname the button shows only the Revised Romanization spelling, such as Gam for 감.
- Check the address order before you use it. The Address tab transcribes by sound and does not translate or reorder. It keeps the Korean order, so you get Seoul-teukbyeolsi Jongno-gu Sejong-daero 175. The page hint says that for mail the English order is the reverse of the Korean one: number and road first, city last. A unit directly after a number is hyphenated, so 종로 2가 becomes Jong-ro 2-ga. If a form wants just Seoul, delete the "-teukbyeolsi" yourself.
Common problems and how to fix them
- It says "Type a Korean name of 2 to 5 Hangul syllables, such as 홍길동."
- The Name box accepts only Hangul syllables, with at most one space after the surname. Latin letters, Chinese characters, digits, a lone consonant such as ㅎ and a single syllable all trigger the message. Retype the name in Hangul. If you only have the Chinese-character form, type its Korean reading.
- The surname and given name are split in the wrong place, for example 황보라 shows Hwangbo and Ra.
- Put a space between the surname and the given name, 황 보라, and the split follows your space. Without a space, the tool treats 남궁, 황보, 제갈, 선우, 독고, 사공 and 서문 at the start as one surname when the name has three or more syllables. With a space, the part before it is the surname.
- Some characters come out unchanged, such as Chinese characters, 漢字, or a lone ㅎ or ㅇ.
- Only complete Hangul syllables (가 to 힣) are converted. Chinese characters, lone jamo, Latin letters, numbers and punctuation are copied through as typed. Replace a lone consonant with a full syllable, and type the Korean reading in place of Chinese characters. Text copied from a macOS file name with split-up letters is joined first, so it converts normally.
- The result differs from a spelling you expected, such as 학여울 giving hagyeoul instead of Hangnyeoul, or 잡혀 giving japhyeo instead of japyeo.
- Two rules need information the tool does not have. Inserted ㄴ sounds, as in 학여울, depend on word boundaries inside a compound, so they are not applied. For ㄱ, ㄷ or ㅂ followed by ㅎ, the tool uses the noun rule, 묵호 gives mukho, because names and places are what people convert; a verb like 잡혀 is therefore written with ph. Tense sounds are never marked, so 압구정 is apgujeong. Edit the few letters by hand if you need the other form.
- The address shows Jong-ro on the Address tab but Jongno on the Text tab.
- That is intended. In an address, a unit such as 로, 길, 구 or 동 is written after a hyphen with no sound change across it, so the road 종로 is Jong-ro and the district 종로구 is Jongno-gu. On the Text tab the word is a plain word, so the sound change applies and 종로 gives jongno. Use the tab that matches what you are writing.
How it compares
The Revised Romanization of Korean is the official system, and this tool follows it, including the sound changes between syllables. A plain letter-by-letter table would write 신라 as sinra and 종로 as jongro; this tool writes Silla and Jongno. It does not do McCune-Reischauer or Yale, it does not translate, and it does not look up official addresses or check a name against your passport. Two cases are simplified: inserted ㄴ in compounds is not applied, and ㅎ after ㄱ, ㄷ or ㅂ always follows the noun rule. For an official document, an official English address or a name already on your passport decides the spelling, not this converter.
Related Tools
Japanese to Romaji Converter
Convert Japanese to romaji: kanji, hiragana and katakana, line by line for song lyrics, in Hepburn romanization. Kanji use a one-time dictionary download; everything runs in your browser. Free.
Roman Numeral Converter
Convert numbers to Roman numerals and back. Standard numerals cover 1–3999, with vinculum notation extending the range to 3,999,999.
OCR — Image to Text Extractor
Extract text from images, screenshots, scanned documents and PDFs (or save a searchable PDF) with PP-OCRv5, running entirely in your browser. Reads Latin, Chinese, Japanese, Korean, Cyrillic, Arabic, Devanagari, Thai, Greek, Tamil and Telugu script.