kongxiag/pdfreader

kongxiag★ 0Python最后同步: 2026-08-19

在 GitHub 打开

README 摘要

pdfreader — an AI literature-reading tool built on pdf-inspector English · 中文版:README.zh.md Convert foreign-language PDF papers into structured Markdown , translate them into Chinese with any OpenAI-compatible LLM (DeepSeek by default; OpenAI / Qwen / GLM / relays also work), and produce bilingual reading reports — making it far faster for both humans and AI to read academic literature. Why pdf-inspector - Smart routing : classifies a PDF as text-based or scanned first, avoiding wasteful OCR on text-based PDFs (slow and costly). - High-quality extraction : implemented in Rust, extremely fast (42 ms for a 10-page test paper), and produces structured Markdown that preserves headings, tables, and layout. - On-demand OCR : the OCR runtime is only initialized for scanned pages; auto mode decides automatically. Installation Quick start Translation model configuration Translation goes through an OpenAI-compatible API . All three values are configurable and default to DeepSeek: Setting Flag Environment variable Model --model (default deepseek-chat ) — Endpoint --base-url (default https://api.deepseek.com/v1 ) — API key --api-key or translate.api key in config LLM API KEY (also reads DEEPSE…

在 GitHub 查看完整 README →
工具/开发aideepseek-harnessdsh-pluginliterature-reviewpdfocrvision

分类