What is DocuTranslate?
DocuTranslate is a free and open-source local-first translation utility designed to bridge the gap between complex document formats and the power of Large Language Models (LLMs).
Unlike other standard web-based translators that often struggle with formatting, this tool is built to handle the heavy lifting of structure preservation.
It supports an impressive array of file types, ranging from standard office documents like Word and Excel to specialized formats like EPUB e-books, subtitles (SRT/ASS), and even raw JSON data. Because it operates locally, it offers a layer of privacy and control that is essential for handling sensitive or proprietary documents.
What sets DocuTranslate apart is its ability to handle the “messy” parts of translation that usually break other tools. It utilizes advanced parsing engines (specifically docling and mineru) to accurately recognize and translate complex elements within PDFs, such as scientific formulas, data tables, and code snippets. This ensures that the layout and meaning remain intact, rather than turning into a jumbled mess.
Moreover, it addresses the common AI issue of inconsistent terminology by automatically generating glossaries, ensuring that specific terms remain aligned throughout the entire document.
Who Can Use It?
This tool is particularly valuable for a wide spectrum of users, from academics and researchers to developers and content creators.
A researcher can translate a foreign scientific paper without losing the integrity of their data tables or mathematical proofs, while a software developer can precisely translate specific values in a JSON file for app localization without breaking the code structure.
Similarly, hobbyists or translators working on novels or subtitles will find the format support indispensable. It is effectively designed for anyone who needs high-fidelity translation where context, structure, and specialized terminology matter just as much as the words themselves.
Key Features
- Comprehensive Document Format Support allows you to translate PDF, Word (DOCX), Excel (XLSX), Markdown, EPUB eBooks, and subtitle files (SRT/ASS) in a single tool.
- Advanced PDF Parsing with Table and Formula Recognition utilizes Docling and Mineru engines to accurately translate academic papers, preserving complex layouts, scientific formulas, and code snippets.
- Precision JSON Translation for Localization offers developer-friendly support for translating specific values using JsonPath syntax, perfect for app and software internationalization.
- Automated Glossary Generation ensures terminology consistency across large projects by creating and adhering to custom term bases during the translation process.
- High-Fidelity Format Preservation maintains the original styling, layout, and structure of Word and Excel documents to eliminate post-translation reformatting.
- High-Performance Asynchronous Processing enables parallel multi-tasking and concurrent translations, designed to handle heavy workloads and batch processing efficiently.
- Multi-Platform AI Integration allows you to connect with various LLM providers and define custom prompts for optimized translation quality and speed.
- Developer-Ready RESTful API and Web UI provides a clean, interactive interface and a robust API for seamless integration into existing workflows and automation pipelines.
- Local Network and Multi-User Capability supports LAN deployment, enabling teams to share translation resources securely within a private network.
- Ultra-Lightweight Portable Deployment delivers a powerful local translation solution in compact packages under 40MB for both Windows and Mac without complex installation.
- Comes with a WebUI Support
- Easy to install using Docker, or from source using Python Package Manager.
License
The project is an open-source project that is released under the MPL-2.0 License.
