[
    {
        "id": "osp-24441",
        "type": "article-journal",
        "title": "Xiaomi-OCR-0 Technical Report",
        "author": [
            {
                "family": "Chen",
                "given": "Xin"
            },
            {
                "family": "Du",
                "given": "Anan"
            },
            {
                "family": "Feng",
                "given": "Feng"
            },
            {
                "family": "Fu",
                "given": "Pei"
            },
            {
                "family": "Luan",
                "given": "Jian"
            },
            {
                "family": "Xu",
                "given": "Longwei"
            },
            {
                "family": "Zhang",
                "given": "Shaojie"
            },
            {
                "family": "Li",
                "given": "Hang"
            },
            {
                "family": "Qu",
                "given": "Heng"
            },
            {
                "family": "Tan",
                "given": "Cheng"
            }
        ],
        "URL": "https://omanscience.com/en/articles/xiaomi-ocr-0-technical-report",
        "language": "en",
        "issued": {
            "date-parts": [
                [
                    2026
                ]
            ]
        },
        "abstract": "Compact OCR-specific vision-language models achieve strong document parsing performance, but often rely on costly supervision and focus primarily on visual-text reconstruction. We introduce Xiaomi-OCR-0, a unified 0.8B model for document parsing and OCR-centric understanding. We build an approximately 170M-sample OCR-centric corpus using an automated data engine that combines expert consensus, render-based verification, and targeted synthesis. Starting from Qwen3.5-0.8B, our progressive training recipe combines Q-Mask-based text anchoring, continued pretraining, and mixed-task reinforcement learning (Mix-RL). Xiaomi-OCR-0 achieves 95.24 on Real5-OmniDocBench, 96.83 on OmniDocBench v1.6, and 87.94 on Wild-OmniDocBench, while reaching an average score of 83.2 across five OCR-oriented VQA benchmarks. Ablations further show that, with sufficient parsing training, OCR-centric understanding supervision provides additional gains for document parsing. Homepage: https://huggingface.co/spaces/SeerRay-Lab/Xiaomi-OCR-0."
    }
]