OpenDataLab ORG

Transforms complex documents like PDFs and Office docs into LLM-ready markdown/JSON for your Agentic workflows.

A Comprehensive Toolkit for High-Quality PDF Content Extraction

DocLayout-YOLO: Enhancing Document Layout Analysis through Diverse Synthetic Data and Global-to-Local Adaptive Perception

[CVPR 2025] A Comprehensive Benchmark for Document Parsing and Evaluation

Open-source multimodal data annotation platform with AI auto-annotation support.

The Open-Source Data Annotation Platform

[ECCV 2026] A diffusion-based framework for document OCR that replaces autoregressive decoding with block-level parallel diffusion decoding.

Agent-native knowledge engine with MCP tools for document indexing, wiki organization, fast retrieval and deep reading across PDF/DOCX/PPTX/Markdown

万卷1.0多模态语料

No description.

No description.

UniMERNet: A Universal Network for Real-World Mathematical Expression Recognition

MinerU-HTML: An SLM-powered HTML main content extractor that outputs clean HTML bodies. Perfect for Deep Research Agents, RAG applications, and training data generation.

No description.

[ACL 2025 Best Theme Paper] This is the official implementation for the paper: "Meta-rater: A Multi-dimensional Data Selection Method for Pre-training Language Models"

[ICLR 2025 Spotlight] The official implementation of the paper “LOKI：A Comprehensive Synthetic Data Detection Benchmark using Large Multimodal Models”

[ICLR 2026] The official implementation of the paper “Earth-Agent: Unlocking the Full Landscape of Earth Observation with Agents”

Data annotation component library --provided as NPM packages

No description.

[NeurIPS 2025 🔥] FakeVLM: Advancing Synthetic Image Detection through Explainable Multimodal Models and Fine-Grained Artifact Analysis

datasets resource

A Python package for interacting with the MinerU Vision-Language Model.

No description.

VHM: Versatile and Honest Vision Language Model for Remote Sensing Image Analysis

Beyond Hallucinations: Enhancing LVLMs through Hallucination-Aware Direct Preference Optimization

(ICCV 2025) OCR Hinders RAG: Evaluating the Cascading Impact of OCR on Retrieval-Augmented Generation

AAAI 2024: Visual Instruction Generation and Correction

[CVPR 2024] 3D Building Reconstruction from Monocular Remote Sensing Images with Multi-level Supervisions

Data browser based on s3. 一个基于 S3 的数据（json / jsonl / parquet / html / md等）可视化工具。👇 Try online.

[ICCV 2025] The official implementation of the paper “Street-to-Satellite Image Synthesis with Diffusion Models and BEV Paradigm”

[ICCV25 Highlight] The official implementation of the paper "LEGION: Learning to Ground and Explain for Synthetic Image Detection"

No description.

ECCV2024_Parrot Captions Teach CLIP to Spot Text

Standardized tool schemas and SDKs that expose Sciverse Open Platform retrieval capabilities to LLM agents.

SDK of OpenDataLab - https://opendatalab.org.cn

MLLM-DataEngine: An Iterative Refinement Approach for MLLM

[ACL 2024 Main Conference] Chinese commonsense benchmark for LLMs

[ICLR 2025] This is the official implementation for the paper: "Large Language Models Meet Symbolic Provers for Logical Reasoning Evaluation"

WanJuan3.0（“万卷·丝路”）一个作为综合性的纯文本语料库，采集了多个国家地区的网络公开信息、文献、专利等资料，数据总规模超1.2TB，Token总数超过300B，处于国际领先水平，首期开源的语料库主要由泰语、俄语、阿拉伯语、韩语和越南语5个子集构成，每个子集的数据规模均超过150GB

Data Set Description Language Specification （新一代人工智能数据集描述语言DSDL）

[AAAI 2025]This repo contains evaluation code for the paper “UrBench: A Comprehensive Benchmark for Evaluating Large Multimodal Models in Multi-View Urban Scenarios”

(CVPR 2026) TRivia: Self-supervised Fine-tuning of Vision-Language Models for Table Recognition

No description.

No description.

[CVPR 2026] SOTA Chemical Reaction Diagram Parsing Framework

No description.

MPB (Miner-PDF-Benchmark) is an end-to-end PDF document comprehension evaluation suite designed for large-scale model data scenarios.

No description.

WebMainBench is a high-precision benchmark for evaluating web main content extraction.

MinerU Training Camp course materials and tutorials

The official implementation of the paper "CrossViewDiff: A Cross-View Diffusion Model for Satellite-to-Street View Synthesis"

WanJuan-CC是以CommonCrawl为基础，经过数据抽取，规则清洗，去重，安全过滤，质量清洗等步骤得到的高质量数据。

No description.

No description.

No description.

No description.

🕶️ A curated list of awesome things related to MinerU

LabelU front-end library

A universal command line tool for compression and decompression

No description.

[AAAI25] Certainty Represented Knowledge Flow for Refusal-Aware Instruction Tuning

[CVPR26 Highlight] The official implementation of the paper "OmniDocLayout: Towards Diverse Document Layout Generation via Coarse-to-Fine LLM Learning"

[NAACL25 findings] Gradient-Driven Refusal-Aware Instruction Tuning for Effective Hallucination Mitigation

No description.

No description.

No description.

A forked repo of the official RDKit library