Is the Pinyin document translator free?

You can try it free, with no account needed, and signing in gives you a larger daily allowance. Pro and Scholar are there for longer documents and regular use.

Can it read a photo of a menu, a sign or a textbook page?

Yes: photographed or scanned pages go through text recognition and then romanization. Printed characters in a plain font read well; decorative and calligraphic fonts, vertical text, small print, glare on a laminated menu and photos taken at an angle all cost accuracy, and a misread character produces a confident, wrong syllable. Shoot straight on, fill the frame with the text and check the output against the picture for any word you will rely on.

Are the tone marks right? What about neutral tones and tone changes?

Tone marks follow the dictionary reading of each character in its context, which is what textbooks and dictionaries print. The known weak points are characters with more than one reading (行, 长, 乐, 了, 得, 地) where the context decides, and the neutral tone, which varies between speakers and regions (the second syllable of 东西 is neutral when it means thing and full when it means east and west). Tone sandhi is normally not written: nǐ hǎo is spelled with two third tones even though it is spoken with the first rising, and 一 and 不 usually keep their citation tone in print. Check any line you will print or teach from against a dictionary.

How accurate is it? What is the difference between Fast and Advanced?

Fast gives a usable pronunciation pass for a menu or a short list. Advanced runs a frontier model and is the choice for textbooks, lyric sheets and anything you will print or teach from: it resolves polyphonic characters from context more reliably, keeps word-level spacing rather than syllable-by-syllable output, and handles names and set phrases more consistently. Neither tier fixes a misread character in a poor photo, so the picture matters as much as the tier.

Does it translate the meaning too? What about Traditional characters, Cantonese and Pinyin back into characters?

No: Pinyin is pronunciation, and the output is the sound of the Mandarin, not an English translation; for the meaning, use the Mandarin Chinese or Traditional Chinese document pages. Simplified and Traditional characters both work as input and give the same Pinyin. Cantonese is a different spoken language with its own romanizations (Jyutping, Yale), so Pinyin for a Hong Kong menu gives Mandarin sounds that a Cantonese speaker does not use. Going the other way, from Pinyin to characters, works too, with the caveat that a syllable without tones (shi, yi, ji) matches dozens of characters and context has to decide.

Will the Pinyin keep my document's layout?

The translator produces a PDF, Word, PowerPoint, Excel or EPUB file that aims to preserve the original layout, including tables, columns and headers. Pinyin is roughly two to three times longer than the characters it replaces, so lines wrap and a tight table cell may need a smaller size. Text size, fonts and line breaks may change, so review the result against the original. Translated PDFs come back as page images; Word files stay editable, which is useful for adding the Pinyin line under the characters yourself.

What file types can I upload?

PDF, DOCX (Word), PPTX (PowerPoint), XLSX (Excel) and EPUB, plus photographed or scanned pages as JPG, PNG or HEIC. Free and Pro accept up to 20MB and 100,000 extracted characters per file; Scholar accepts up to 50MB and 1,000,000 characters.

Pinyin PDF & Document Translator — Chinese Documents, Photos & Scans into Pinyin

Romanize a Chinese PDF, Word file, menu, sign, textbook page or photo into Hanyu Pinyin with tone marks while keeping the document's layout. Upload the file, keep Pinyin as the target, and download the result. Pinyin is pronunciation, not meaning: for an English translation, use the Mandarin Chinese document page.

Loading credits…

Translate into

Drag and drop a file here, or click to browse

PDF
DOCX
PPTX
XLSX
EPUB
JPG
PNG
TXT

See “Keep formatting” in action

Real examples of the “Keep formatting” option: original on the left, translated document on the right. Click either page to enlarge.

1 / 6

A civil complaint filed in Mexico City, Spanish to English. Caption block, numbered pleadings, the evidence table and the signature block all keep their places. — Original — A civil complaint filed in Mexico City, Spanish to English. Caption block, numbered pleadings, the evidence table and the signature block all keep their places.

An Italian birth certificate, photographed rather than scanned. The framed masthead, the field grid, the marginal annotations and the round seal all hold their positions. — Original — An Italian birth certificate, photographed rather than scanned. The framed masthead, the field grid, the marginal annotations and the round seal all hold their positions.

A handwritten archaeological field notebook from 1974. The engine reads the hand and keeps the ruled inventory table intact. — Original — A handwritten archaeological field notebook from 1974. The engine reads the hand and keeps the ruled inventory table intact.

A handwritten archaeological field notebook from 1974. The engine reads the hand and keeps the ruled inventory table intact. — Translation — A handwritten archaeological field notebook from 1974. The engine reads the hand and keeps the ruled inventory table intact.

A German hospital discharge summary. Diagnoses with ICD codes, lab values and the medication table survive the translation unchanged. — Original — A German hospital discharge summary. Diagnoses with ICD codes, lab values and the medication table survive the translation unchanged.

A German hospital discharge summary. Diagnoses with ICD codes, lab values and the medication table survive the translation unchanged. — Translation — A German hospital discharge summary. Diagnoses with ICD codes, lab values and the medication table survive the translation unchanged.

A 1747 parish baptism register in Latin. Ruled columns, godparents and marginal notes stay aligned, so the genealogy still reads as a register. — Original — A 1747 parish baptism register in Latin. Ruled columns, godparents and marginal notes stay aligned, so the genealogy still reads as a register.

A 1747 parish baptism register in Latin. Ruled columns, godparents and marginal notes stay aligned, so the genealogy still reads as a register. — Translation — A 1747 parish baptism register in Latin. Ruled columns, godparents and marginal notes stay aligned, so the genealogy still reads as a register.

A scanned 17th-century Italian index of sermons. Two columns of dense small print, translated in place. — Original — A scanned 17th-century Italian index of sermons. Two columns of dense small print, translated in place.

A scanned 17th-century Italian index of sermons. Two columns of dense small print, translated in place. — Translation — A scanned 17th-century Italian index of sermons. Two columns of dense small print, translated in place.

How Pinyin documents come out

Upload a photo of a menu, a sign, a textbook page, a worksheet, a lyric sheet or a name list, or a Chinese PDF or Word file, and get the text back in Hanyu Pinyin with tone marks, laid out as the original. Pinyin is the pronunciation of Mandarin in Latin letters, not a translation: nǐ hǎo tells you how to say 你好, not what it means. For the meaning, use the Mandarin Chinese or Traditional Chinese document pages. Simplified and Traditional characters both work as input, and Pinyin can also be turned back into characters.

Hanyu Pinyin is the official romanization of Standard Mandarin, adopted in mainland China in 1958 and now the international standard for Chinese names and teaching. Each syllable takes a tone mark on its main vowel: first tone mā, second má, third mǎ, fourth mà, and the neutral tone is left unmarked (māma). The vowel ü keeps its dots after l and n (lǜ, nǚ) and loses them after j, q, x and y. Proper Pinyin is written by word, not by character: Běijīng, not Běi Jīng; surnames capitalized and given names run together (Wáng Xiǎomíng). Two features make Chinese to Pinyin harder than it looks. Many common characters have more than one reading (多音字) and only the context decides it: 行 is háng in 银行 (bank) and xíng in 行走 (walk), and 长 is cháng (long) or zhǎng (to grow). And the tones written are usually the dictionary's citation tones even where speech changes them: two third tones in a row are spoken with the first rising (nǐ hǎo sounds like ní hǎo), and 一 and 不 shift tone before certain syllables; some textbooks mark the change, and either convention is fine if it is consistent. Photos go through text recognition first, so the Pinyin is only as good as the character reading, and stylized or calligraphic signs are the hardest input.

Translate your document and keep its structure

The translator keeps your file’s structure while translating the text in context. The downloaded format depends on the source file.

PDFs and scanned pages: The translated PDF follows the page visually, with translated pages rendered as images. Selectable PDF text and identical fonts are not guaranteed.
Word documents and PowerPoint presentations: Start with DOCX or PPTX when you have it. Native paragraphs, table text and slide text can stay editable in the corresponding file format. Translated text inside pictures remains image content; review text wrapping and font changes in your editor.
Excel workbooks and EPUB books: Use the original XLSX or unprotected EPUB for a translated workbook or ebook. Review formulas, numbers and text cells in a workbook; check chapter order, navigation and text flow in your ebook reader.

What people translate to and from Pinyin

Menus, signs and shop fronts

The most common upload is a phone photo: a restaurant menu, a street sign, a shop front, a station board, a product label. Pinyin lets you say the dish or the place out loud (宫保鸡丁 is gōngbǎo jīdīng, 出口 is chūkǒu) and match it to what a speaker says back, which is what a traveller or a learner ordering food wants. A menu laid out in columns keeps its columns. Calligraphic fonts, vertical text and glossy laminated menus photographed at an angle are the hard cases, so shoot straight on and crop to the text. For what the dish actually is, translate the same photo into English instead.

Textbooks, worksheets and character lists

Learners add Pinyin to a reading passage, a vocabulary list, an HSK practice page or a children's reader, and teachers make a Pinyin key for a worksheet whose characters have none. The layout is kept, so a two-column vocabulary table stays a table and a numbered exercise stays numbered. Word-level spacing is what a textbook expects, so check that compounds came out as words rather than as syllable strings, and compare any character with several readings against the book's own gloss. Textbooks from Taiwan annotate with Bopomofo (Zhuyin) rather than Pinyin; the Bopomofo text page covers that system, and Traditional-character input works here just the same.

Song lyrics, poems and karaoke sheets

Lyric sheets for singing along, a Tang poem with the pronunciation under each line, a hymn or a children's rhyme: line-by-line Pinyin is what these need, and the line breaks of the original are kept. Two cautions. Sung Chinese ignores tones, so tone marks on a lyric sheet are for reading rather than for the tune, and neutral tones in lyrics are often given their full tone. And classical poems read in modern Mandarin no longer rhyme where they once did, so a Pinyin rendering of a Tang poem is its modern pronunciation, not the sound the poet heard, which matters if you are studying the metre.

Names, addresses and forms

A class roster, a staff list, an address for a parcel, a family register, a business card: Pinyin gives a pronounceable form of each name, with the surname first and capitalized and the given name as one word (Wáng Xiǎomíng). Two cautions apply. A person's own romanization takes precedence, and many people do not use Pinyin: Hong Kong names are usually spelled from Cantonese (Lee, Chan, Wong) and Taiwanese names often follow older systems (Chang, Tsai, Hsu), spellings that cannot be recovered from the characters alone. And a name's characters may have several readings, so a roster is where the polyphonic-character problem shows up most, and a quick check with the person is worth more than any tool.

What to watch for in Pinyin document translation

The first trap is polyphonic characters: 行, 长, 乐, 了, 重, 得 and dozens more change reading with meaning, and a wrong pick is fluent and invisible unless you know the word. The second is segmentation: Pinyin is spelled by word (Zhōnghuá Rénmín Gònghéguó), and output that spaces every syllable or runs a whole phrase together is wrong by the orthography even when every syllable is right. Tone marks sit on the main vowel by rule (a or e first, then o in ou, otherwise the last vowel: iù, uì, hǎo), the apostrophe separates syllables that would otherwise merge (Xī'ān is two syllables, xiān is one), ü is written u after j, q, x and y and must keep its dots after l and n (lǜ, not lu), and the Beijing r ending is written with a final r (huār). Tone sandhi is a matter of convention, so decide whether your document writes citation tones or spoken tones and keep it consistent. Numbers, names and code-like strings are the parts a translation engine may leave in characters, so check headings, name lists and anything that looks like a label. Readings differ between mainland and Taiwan for some characters (垃圾 lājī or lèsè), and Classical Chinese read in Mandarin is a modern pronunciation of an old text. Finally, a romanized document is still Chinese: readers who want the meaning need a translation, not Pinyin.

Pinyin document translation FAQ

Is the Pinyin document translator free?: You can try it free, with no account needed, and signing in gives you a larger daily allowance. Pro and Scholar are there for longer documents and regular use.
Can it read a photo of a menu, a sign or a textbook page?: Yes: photographed or scanned pages go through text recognition and then romanization. Printed characters in a plain font read well; decorative and calligraphic fonts, vertical text, small print, glare on a laminated menu and photos taken at an angle all cost accuracy, and a misread character produces a confident, wrong syllable. Shoot straight on, fill the frame with the text and check the output against the picture for any word you will rely on.
Are the tone marks right? What about neutral tones and tone changes?: Tone marks follow the dictionary reading of each character in its context, which is what textbooks and dictionaries print. The known weak points are characters with more than one reading (行, 长, 乐, 了, 得, 地) where the context decides, and the neutral tone, which varies between speakers and regions (the second syllable of 东西 is neutral when it means thing and full when it means east and west). Tone sandhi is normally not written: nǐ hǎo is spelled with two third tones even though it is spoken with the first rising, and 一 and 不 usually keep their citation tone in print. Check any line you will print or teach from against a dictionary.
How accurate is it? What is the difference between Fast and Advanced?: Fast gives a usable pronunciation pass for a menu or a short list. Advanced runs a frontier model and is the choice for textbooks, lyric sheets and anything you will print or teach from: it resolves polyphonic characters from context more reliably, keeps word-level spacing rather than syllable-by-syllable output, and handles names and set phrases more consistently. Neither tier fixes a misread character in a poor photo, so the picture matters as much as the tier.
Does it translate the meaning too? What about Traditional characters, Cantonese and Pinyin back into characters?: No: Pinyin is pronunciation, and the output is the sound of the Mandarin, not an English translation; for the meaning, use the Mandarin Chinese or Traditional Chinese document pages. Simplified and Traditional characters both work as input and give the same Pinyin. Cantonese is a different spoken language with its own romanizations (Jyutping, Yale), so Pinyin for a Hong Kong menu gives Mandarin sounds that a Cantonese speaker does not use. Going the other way, from Pinyin to characters, works too, with the caveat that a syllable without tones (shi, yi, ji) matches dozens of characters and context has to decide.
Will the Pinyin keep my document's layout?: The translator produces a PDF, Word, PowerPoint, Excel or EPUB file that aims to preserve the original layout, including tables, columns and headers. Pinyin is roughly two to three times longer than the characters it replaces, so lines wrap and a tight table cell may need a smaller size. Text size, fonts and line breaks may change, so review the result against the original. Translated PDFs come back as page images; Word files stay editable, which is useful for adding the Pinyin line under the characters yourself.
What file types can I upload?: PDF, DOCX (Word), PPTX (PowerPoint), XLSX (Excel) and EPUB, plus photographed or scanned pages as JPG, PNG or HEIC. Free and Pro accept up to 20MB and 100,000 extracted characters per file; Scholar accepts up to 50MB and 1,000,000 characters.

See “Keep formatting” in action

Real examples of the “Keep formatting” option: original on the left, translated document on the right. Click either page to enlarge.

1 / 6

Translate documents in other languages

Malay (Jawi) documents Malay documents Akkadian documents Old English documents Arabic documents Aramaic documents Moroccan Arabic documents Egyptian Arabic documents Bavarian documents Bengali documents Pinyin text translator All document translation