Translate a PDF to English, with the original beside it
Extract a PDF's text, automatically detect its language, and review editable English next to each original segment. Paperly supports cloud translation for languages such as Spanish, Japanese and French.
Open the English translator ↗ Start by converting a PDF
Updated . Free beta, no account required.
How to translate a PDF into English
- Extract the text locally. Open the converter, choose your PDF and press Convert. Files can be up to 50 MB each, with up to 8 selected at once. Review the extracted text before translating. Use a PDF page range to work on selected pages.
- Send the result to the translator. Choose “Translate to English” on the result card. It copies the current edited Markdown into the translation input. You can also paste text directly or use text extracted from an Office document.
- Choose a translation method. Cloud mode uses automatic source-language detection by default. You can select a source language manually if the detector gets short text wrong. Tick the permission box before sending text through Paperly to Alibaba Cloud.
- Compare and edit. Read each original segment beside the English field, check the detected language code, and edit wording directly. Your original stays available. Long text is processed in segments, and you can load more comparison rows as needed.
- Download or continue exporting. Save English Markdown, download the bilingual ZIP, or send your edited English to the document export tools for Word, HTML, EPUB and browser Print / Save as PDF.
Spanish and Japanese PDFs to English
A PDF with selectable Spanish or Japanese text can follow the same extraction-and-translation workflow. These short examples were checked with the cloud translator; they illustrate the text comparison rather than a promise about every document's quality.
| Original language | Original text | English result |
|---|---|---|
| Spanish (es) | Hola, lea este documento con atención. | Hello, please read this document carefully. |
| Japanese (ja) | この文書をよく読んでください。 | Please read this document carefully. |
| French (fr) | Veuillez lire ce document attentivement. | Please read this document carefully. |
English input is kept in English. Automatic detection is not a guarantee for every language, dialect or mixed-language sentence. If the original is misidentified, try the manual source-language selector or use a longer coherent passage. Supported language combinations depend on the translation service.
Original and translated text, side by side
The comparison pairs extracted text segments with editable English. This is useful when checking names, dates, quantities or instructions against the source. It is not an overlay on the original PDF, a page-image viewer or a recreation of the original page design. Multi-column reading order and table extraction should be checked before translation, because a translator cannot reliably repair missing source content.
What is in the bilingual download?
| File or export | What it contains |
|---|---|
| original.md | The extracted or pasted source text used for this translation. |
| english.md | The English text, including your edits. |
| comparison.html | A standalone original/English table. Extract the ZIP and open this file to read both columns or print it to PDF. |
| Word and other document exports | Use “Export English as document” to fill the existing export editor. Formatting is basic and may change. |
Long documents without a character quota
Paperly does not impose a translation document character cap or daily character quota. It splits long text into requests automatically and displays comparison rows in batches. Loading more rows changes what is visible, not what is translated or exported. Keep the tab open until translation finishes; you can cancel a running job. Partial translations cannot be exported as a completed English document.
This does not remove the converter's 50 MB per-file selection limit or the practical limits of browser memory, device performance and cloud service availability. A long scan also takes time to OCR. Work on a smaller page range if your device struggles. Safari and iPhone have not been verified on real devices.
Cloud translation and local processing are different
PDF conversion and OCR happen in the browser. Cloud translation is optional: when you click Translate to English, extracted text travels through Paperly's server to Alibaba Cloud Machine Translation in the Hangzhou region. The original PDF file is not uploaded. Paperly does not store translation text or results or maintain IP-based character quota counters; the cloud provider handles the text under its own policies. Read the privacy notice before sharing confidential material.
Scans, tables and layout: what to check
Multilingual text translation does not mean multilingual scan OCR. Paperly's current OCR models read English and Simplified Chinese. A scanned Japanese, Spanish or other-language PDF needs a suitable OCR step elsewhere before you paste readable text into the translator. See the scanned PDF OCR guide for the difference between a text PDF and an image-only scan.
Translation leaves Markdown tables, fenced code, metadata and image references unchanged. Ordinary PDF illustrations are not extracted, and complex tables may lose structure. The result is an editable text translation, not a certified translation or a pixel-matched translated PDF. Check important figures and wording; do not rely on unreviewed machine output for consequential decisions.
Questions about PDF translation
Can Paperly automatically detect the original language?
Yes. Cloud translation defaults to automatic source-language detection and translates supported languages to English. Spanish, Japanese, French and English have been tested. Short or mixed-language text can be misidentified; select the source language manually if needed.
Can I read the original and English translation side by side?
Yes. Each extracted text segment is shown beside an editable English field. The comparison is text-based; it does not reproduce the original PDF page layout.
What can I download after translation?
Download English Markdown or a bilingual ZIP containing original.md, english.md and comparison.html. Open comparison.html to read both columns or print it to PDF. You can also send the edited English to the Word, HTML, EPUB and other export tools.
Is there a document character limit or daily quota?
Paperly has no translation document character limit or daily character quota. Long text is processed in segments and comparison rows load in batches. Browser memory, cloud service availability and the converter's 50 MB per-file upload-selection limit still affect large files.
Does cloud translation send my PDF to a server?
File conversion and OCR run locally. When you click Translate to English, extracted text goes through Paperly to Alibaba Cloud Machine Translation; the original PDF file is not uploaded.
Can Paperly translate a scanned Japanese or Spanish PDF?
The translation service supports Japanese and Spanish text, but Paperly's current scan OCR models support English and Simplified Chinese only. Other-language scans need readable text from a suitable OCR tool before you paste or import the text.
Does it preserve every table, formula and image?
No. Complex extraction can lose structure. Translation leaves Markdown tables, code blocks, metadata and image references unchanged, and ordinary PDF illustrations are not extracted. Review names, numbers and important wording against the original.
Related: PDF to Markdown · English and Chinese OCR · Free beta features