Fast, spatial PDF parsing in Rust — column-aware text extraction, optional OCR, and format conversion. Competitive with LiteParse on real-world documents.

The first public benchmark for document translation with layout preservation. LTB-100 = chrF + Layout IoU + Reading-order Kendall tau.

An open-source desktop app & Python API/CLI for high-fidelity, layout-preserving PDF translation. Features interactive split-pane viewing with synchronized scrolling, bilingual hover peek, and a resilient, multi-provider LLM engine (Gemini, OpenAI, Claude, Ollama) that preserves complex academic document geometry.

Claude/Codex Skill：英文 PDF 版式保真中文翻译，面向研报、公告、合同和白皮书，尽量保留表格、图表和页面结构。

🚀 A premium, lightweight, layout-preserving desktop PDF editor for Windows. Edit text inline (behaves like Microsoft Word), extract fonts, replace/resize images, rotate/reorder pages, and run Tesseract OCR on scanned documents. Built with Electron, React, TypeScript, Tailwind CSS, and a pure-Python FastAPI processing engine.

Agent skill for translating PDFs between 20+ languages with full layout preservation. Uses PDFMathTranslate + Google Translate. Zero API keys needed.

Desktop PDF translator that keeps the original layout. 34 languages, offline or online, with OCR for scanned forms.

Claude Code skill: translate PDFs preserving layout | Claude Code PDF翻译技能（保留排版）| PDF翻訳スキル（レイアウト保持）