pdf2zh-desktop offers a user-friendly, out-of-the-box desktop application for translating academic PDFs, addressing the complexity of traditional Python-based tools. It targets researchers and power users, providing a seamless experience that preserves original document formatting, integrates with Zotero, and supports numerous AI translation engines.

How It Works

This project enhances the PDFMathTranslate engine with a native graphical user interface (GUI) for Windows and macOS, eliminating the need for Python or environment setup. It employs intelligent chunking and memory management for handling large documents, offers built-in OCR for scanned PDFs, and features a dedicated pipeline for accurate table translation. The architecture prioritizes ease of use through drag-and-drop batch processing and a continuous PDF preview.

Quick Start & Requirements

• Installation: Download, unzip, and run the executable (.bat for Windows, .app for macOS). No Python or pip installation required.
• Platforms: Windows 10/11 (64-bit), macOS 13.0+.
• Prerequisites: Minimum 4GB RAM (8GB recommended for large documents), ~750MB disk space. Online translation services require internet connectivity.
• Resources: Official releases available at https://github.com/AaronGIG/pdf2zh-desktop/releases.

Highlighted Details

• AI Translation & Customization: Integrates over 20 AI translation engines (including DeepSeek, GPT-4o, Gemini, Ollama) with support for OpenAI-compatible APIs, custom prompts, and termbases.
• Advanced PDF Handling: Supports batch translation via drag-and-drop, OCR for scanned documents (RapidOCR), a dedicated table translation pipeline, and handles documents exceeding 1000 pages with chunking and breakpoint resume.
• Zotero Integration: Offers deep integration via a connector plugin, enabling drag-and-drop from Zotero and automatic attachment of translated files back into the reference manager.
• Output Formats: Generates translated-only, bilingual side-by-side, or original-left/translated-right PDF outputs.

Maintenance & Community

The project is primarily developed by @AaronGIG (CAS), with contributions noted from Claude (Anthropic AI) for desktop enhancements. A QQ user group (1094195179) is available for community interaction.

Licensing & Compatibility

The specific license for pdf2zh-desktop is not explicitly stated in the provided README. Compatibility for commercial use is not detailed.

Limitations & Caveats

macOS users may need to bypass security prompts for initial app execution. Windows 11 24H2 users should use the .bat file due to VBScript deprecation. The project's license requires clarification for adoption decisions.