---
title: opendatalab/MinerU2.5-Pro-2604-1.2B
description: "MinerU2.5-Pro: Pushing the Limits of Data-Centric Document Parsing at Scale. qwen2_vl, 1.2B parameters."
canonical_url: https://superlinked.com/models/opendatalab-mineru2-5-pro-2604-1-2b
last_updated: 2026-06-29
---

# opendatalab/MinerU2.5-Pro-2604-1.2B

MinerU2.5-Pro: Pushing the Limits of Data-Centric Document Parsing at Scale

Source: [opendatalab/MinerU2.5-Pro-2604-1.2B on HuggingFace](https://huggingface.co/opendatalab/MinerU2.5-Pro-2604-1.2B)

## Overview

| Field | Value |
|-------|-------|
| Architecture | qwen2_vl |
| Parameters | 1.2B |
| Tasks | Extract |
| Outputs | Entities |
| License | apache-2.0 |
| Inputs | image |
| Languages | zh, en |

## Benchmarks

### olmOCR-Bench

Domain: general · Task: ocr · Language: en

Document text extraction accuracy across arxiv math, old scans, multi-column layouts, and tables

Corpus: 1,403 · Queries: 1,403

**Quality:** accuracy: 0.5910

[Reference](https://huggingface.co/datasets/allenai/olmOCR-bench)