BabelDOC

BabelDOC is an open-source document translation tool that translates PDF files, especially scientific papers, while preserving their original layout. Published on GitHub by the FunStory AI team, it has gathered more than 9,700 stars and is under active development, with updates continuing through 2026.

Overview

Translating a PDF is notoriously messy: two-column layouts, equations, tables and figures usually fall apart. BabelDOC parses the document structure first, identifies text blocks and their formatting, protects math formulas and references from translation, and then rebuilds a polished PDF in the target language. It can produce dual-language documents that show the original and translation side by side or interleaved, which makes it a favorite companion for readers of foreign-language research. It also powers translation features in the wider PDFMathTranslate ecosystem.

Key Features

• Layout-preserving translation for academic PDFs, including multi-column papers
• Formula and citation protection, so equations and references are never machine-translated
• Dual-language output modes for reading original and translation together
• Works with many translation backends, including OpenAI-compatible APIs, DeepL, Google Translate and local Ollama models
• Command-line and Python API operation, suitable for batch processing large document collections
• Glossary support to keep technical terminology consistent across a document

💳 Pricing

BabelDOC is free and open source under the AGPL-3.0 license. The software costs nothing; you only pay if you choose a commercial translation API as the backend, and free backends such as Google Translate or local models keep it entirely cost-free.

Who Is It For

BabelDOC is ideal for researchers, graduate students and engineers who regularly read papers, manuals or books published in other languages and want a faithful, readable PDF rather than a wall of extracted text. Localization teams processing technical documentation will also appreciate its batch-friendly design.