paddleocr-vl 开源项目推荐（29 个）

PaddlePaddle /PaddleOCR

将任意 PDF 或图像文档转换为面向 AI 的结构化数据。这是一款强大、轻量的 OCR 工具包，弥合图像/PDF 与 LLM 之间的差距。支持 100 多种语言。

magicyuan876 /mineru-tianshu

天枢 - 企业级 AI 一站式数据预处理平台 | PDF/Office转Markdown | 支持MCP协议AI助手集成 | Vue3+FastAPI全栈方案 | 文档解析 | 多模态信息提取

JhuoW /OpenMathpix

An open-source alternative to Mathpix Snip Tool. Instantly extract LaTeX, math equations, and text from images and screenshots. | 开源 Mathpix 替代工具，支持公式截图转 LaTeX

CHEN010325 /paddleocr-local

本地优先的文档解析工作台，面向五种 OCR 模型：支持 PDF/Office 转 Markdown，并提供 WebUI、模型切换、CLI 和 Apple Silicon 支持。

ablanco1950 /LicensePlate_Yolov8_Filters_PaddleOCR

通过Yolov8自动检测任何格式的车牌号码，通过滤波管线和Paddleocr作为OCR实现识别

Edgaras0x4E /paddleocr-pdf-api

A self-hosted PDF OCR API that converts scanned documents to markdown. Powered by PaddleOCR-VL, runs on GPU via Docker.

mlx-community /paddleocr-vl.swift

Native Swift + MLX port of Paddle0CR-VL, a 0.9B doc parsing VLM for OCR, tables, formulas, and charts on Apple Silicon

EricRollei /PDF-Tools

Advanced PDF processing, OCR, and AI vision analysis nodes for ComfyUI. Extract images from PDFs, perform multilingual OCR with Surya, detect objects with Florence-2, and analyze document layouts.

egore4606 /paddle-ocr-ui

Local-first web UI for PaddleOCR-VL: OCR PDFs and images with Docker, live logs, previews, and ZIP exports.

jinhyeonseo01 /FastFileOCR

Rust中的图像和PDF快速准确的OCR，由PaddleOCR-DL 1.6提供支持。

Romain-Jochum /PaddleOCR-VL_MVP

Minimal script to run PaddleOCR-VL on a PDF and output a MD file with images

GravityPoet /Open-LLM-OCR

Bob 终极 OCR：支持本地隐私模型 + 大模型API云端 OCR 高精度，一键切换。快、免费、隐私好。识别能力更强，可选更多模型

albaNnaksqr /OcrParser

Production-oriented PDF OCR parsing platform for large batch, distributed control/worker workflows with shared storage and OpenAI-compatible model services.

ZengLiangYi /paddleocr-vl-document-parser-demo

零依赖的百度智能云 PaddleOCR-VL 文档解析 API 评估 Demo

945144024 /lsdyna-manual-to-markdown

Convert user-supplied LS-DYNA Keyword and Theory Manual PDFs into structured, traceable Markdown documents.

AIwork4me /omnidocbench-amd-windows

One-command OmniDocBench v1.6 evaluation setup on AMD Windows (ROCm/HIP). Model-agnostic: swap any document parsing model via adapters. PaddleOCR-VL-1.6 as reference. CDM formula scoring included.

ConcatenateLine /paddle-vlm-editor

PaddleVLM Editor is a local, Docker-based application that extracts structured content from PDFs and images using PaddleOCR-VL's Vision-Language Model, then lets you review and edit the results in a rich-text block editor — all without sending data to any cloud service.

valve530 /OCRDflash

OCR-DFlash is a research runner for accelerating PDF document parsing with PaddleOCR-VL, PP-DocLayoutV3, PDF native text drafts, and bounded LLMA-style token verification.

cobaltna /local-ocr

Give any AI agent offline eyes: batch-OCR scanned PDFs & images with PaddleOCR-VL-1.5 (GGUF) on LM Studio — works with text-only LLMs, 100% local & private, outputs text / LaTeX / Markdown tables. 给 AI Agent 装上离线眼睛：纯本地批量 OCR，公式转 LaTeX、表格转 Markdown。

generalMG /DocumentParser

Python data pipeline for arXiv metadata: SQLAlchemy + Alembic schema, PostgreSQL storage, PDF download tracking, and optional PaddleOCR processing.