---

Unlimited OCR Works

Pioneering the OCR-3.0 Era: From Long- to Unlimited-Horizon Parsing.

Release

• [2026/11/02] 📄 The paper of Unlimited OCR 2 is now available on arXiv.
• [2026/10/28] Thanks to the ms-swift community for supporting training of Unlimited OCR 2.
• [2026/10/27] 📊 We release MarathonDocBench, a new benchmark for ultra-long document parsing.
• [2026/10/27] 🤝 Thanks to the ModelScope community for their support. Our model is now available at ModelScope.
• [2026/10/27] 🤝 Thanks to the Baidu Cloud team for their support. Our model is now available on Baidu Cloud.
• [2026/10/27] 🤝 Thanks to AK for creating a new demo for us. Try it at Unlimited-OCR-2.
• [2026/10/27] 🚀 We release Unlimited OCR 2, achieving unlimited-horizon parsing with SR-SWA.
• [2026/07/21] 🤝 Thanks to the ms-swift community for their support, our model now supports training with ms-swift.
• [2026/07/03] 🤝 Thanks to the Baidu Cloud team for their support. Our model is now available on Baidu Cloud.
• [2026/06/28] 🤝 Thanks to the vLLM community and Tianyu Guo for their support, our model now supports vLLM inference.
• [2026/06/24] 🤝 Thanks to AK for creating a demo for us. It is now available at Hugging Face Spaces.
• [2026/06/23] 📄 Our paper is now available on arXiv.
• [2026/06/23] 🤝 Thanks to the ModelScope community for their support. Our model is now available at ModelScope.
• [2026/06/22] 🚀 We present Unlimited-OCR, aiming to push Deepseek-OCR one step further.

What's New in 2

• Sparse Reference Sliding Window Attention (SR-SWA): Compresses text-visual attention interactions from ultra-long sequences down to a fixed-size set of working references (e.g., 512 tokens). This bounds computational complexity, keeping per-step decoding latency and memory footprint nearly constant regardless of document length.
• From Long- to Unlimited-Horizon Parsing: Powered by SR-SWA, Unlimited-OCR 2 achieves true, unbounded document parsing within a finite context window—effectively eliminating quadratic attention scaling and memory bottlenecks as documents grow.
• Ultra-Long Document Parsing SOTA: Achieves state-of-the-art performance across both dozens-of-pages benchmarks (e.g., MPDocBench) and extreme hundred-page benchmarks (e.g., MarathonDocBench).

Inference

Unlimited OCR 2 supports Transformers (native, for quick tests) and SGLang (recommended for serving).

Transformers

Tested with Python 3.10, torch==2.13.0, transformers==4.52.1, huggingface-hub==0.36.0, tokenizers==0.21.4, on an NVIDIA A800-80GB:

pip install torch==2.13.0 torchvision pillow transformers==4.52.1 huggingface-hub==0.36.0 \
  tokenizers==0.21.4 addict==2.4.0 einops==0.8.2 easydict==1.13

Load the model once:

import torch
from transformers import AutoModel, AutoTokenizer

model_name = "baidu/Unlimited-OCR-2"
tokenizer = AutoTokenizer.from_pretrained(model_name, trust_remote_code=True)
model = AutoModel.from_pretrained(model_name, trust_remote_code=True, use_safetensors=True,
                                  torch_dtype=torch.bfloat16).eval().cuda()

Multi-page inference (infer_multi). All pages of a document are parsed in one request, so the model can use cross-page context. Pass the page images in page order:

pages = [f"your_doc/page_{i}.jpg" for i in range(1, 8)]

text, num_tokens = model.infer_multi(tokenizer, image_files=pages,
                                     output_path="output/multi", save_results=True)

The prompt is built automatically as "<PAGE><image>" * num_pages + "Multi page parsing.", and each page is encoded as one global view (base mode). The output contains one <PAGE> per input page, in the same order as image_files.

Single-page inference (infer). A single page is encoded with dynamic tiling (gundam mode) by default:

text = model.infer(tokenizer, image_file="your_doc/page_1.jpg",
                   output_path="output/single", save_results=True)

SGLang

Set up the environment (uv-managed virtualenv, Python 3.10) and install the customized SGLang wheel sglang-0.5.20.dev13+g75b9b06b02-py3-none-any.whl, plus PyMuPDF for PDF-to-image conversion:

uv venv --python 3.10
source .venv/bin/activate

uv pip install wheel/sglang-0.5.20.dev13+g75b9b06b02-py3-none-any.whl
uv pip install pymupdf==1.27.2.2
uv pip install fastokens==0.3.2

Start the SGLang server:

# Page-window scheduling for very long documents (the KV tail of the previous window is reused).
export SGLANG_UOCR_EANABLE_DOC_LIMIT=1          # 1 = enable, 0 = whole document in one window
export SGLANG_UOCR_DOC_GUNDAM_PAGE_WINDOW=10    # pages per window (gundam mode)
export SGLANG_UOCR_DOC_BASE_PAGE_WINDOW=10      # pages per window (base mode)
export SGLANG_UOCR_MAX_PAGE_TOKENS=20000        # per-page output cap; must be > 0
# Chunk cap for the SAM attention bias. The built-in default (512) allows a single
# ~462 MiB allocation and can OOM; 64 is equivalent with a lower peak.
export SGLANG_OPT_UOCR_SAM_REL_POS_BIAS_MB=64
# Pass large crop tensors through /dev/shm instead of pickling them into the
# tokenizer -> scheduler frame.
export SGLANG_UOCR_SHM_WRAP_MODEL_DATA=1
export SGLANG_UOCR_SHM_WRAP_MIN_MB=1
export SGLANG_VLM_CACHE_SIZE_MB=0
# Reduces fragmentation failures on the large contiguous allocations.
export PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True

python -m sglang.launch_server \
    --model-path baidu/Unlimited-OCR-2 \
    --served-model-name Unlimited-OCR-2 \
    --chat-template unlimited-ocr \
    --tokenizer-backend fastokens \
    --tokenizer-worker-num 8 \
    --mem-fraction-static 0.7 \
    --max-running-requests 64 \
    --max-prefill-tokens 2048 \
    --enable-custom-logit-processor \
    --enable-streaming-session \
    --allow-auto-truncate \
    --disable-radix-cache \
    --mm-preprocess-cache-size-mb 0 \
    --cuda-graph-backend-decode full --cuda-graph-max-bs-decode 64 \
    --cuda-graph-backend-prefill full \
    --cuda-graph-max-bs-prefill 16384 \
    --cuda-graph-bs-prefill 1024 2048 \
    --cuda-graph-config '{"prefill": {"backend": "full", "full_prefill_max_req": 64}}' \
    --host 0.0.0.0 \
    --port 10000

Notes on the knobs that matter for memory:

• --mem-fraction-static is the share reserved for weights + KV. 0.75 leaves too little headroom at 32+ concurrent requests; 0.7 is stable.
• --max-prefill-tokens drives a temporary tensor of roughly [tokens, context/32], so memory grows linearly with it: 4096 -> 2048 halves that peak.
• The largest --cuda-graph-bs-prefill bucket must equal --max-prefill-tokens. A smaller one makes oversized batches fall back to eager; a larger one is never used and only wastes capture memory.
• --cuda-graph-config's full_prefill_max_req is the request-slot count for the full-prefill CUDA graph. The default (8) makes 16/32 concurrency fall back to eager.
• A smaller page window lowers peak memory but hurts cross-page merging, since blocks that continue across a window boundary lose their context.

To parse a whole document in one window instead (SGLANG_UOCR_EANABLE_DOC_LIMIT=0).

Send requests to the OpenAI-compatible API:

import base64
import json
import os
import tempfile

import fitz
import requests
from sglang.srt.sampling.custom_logit_processor import DeepseekOCRNoRepeatNGramLogitProcessor

server_url = "http://127.0.0.1:10000"
NGRAM_PROCESSOR = DeepseekOCRNoRepeatNGramLogitProcessor.to_str()

session = requests.Session()
session.trust_env = False


def pdf_to_images(pdf_path, dpi=300):
    doc = fitz.open(pdf_path)
    tmp_dir = tempfile.mkdtemp(prefix="pdf_ocr_")
    mat = fitz.Matrix(dpi / 72, dpi / 72)
    image_paths = []
    for i, page in enumerate(doc):
        image_path = os.path.join(tmp_dir, f"page_{i + 1:04d}.png")
        page.get_pixmap(matrix=mat).save(image_path)
        image_paths.append(image_path)
    doc.close()
    return image_paths


def encode_image(image_path):
    ext = os.path.splitext(image_path)[1].lower()
    mime = "image/jpeg" if ext in (".jpg", ".jpeg") else f"image/{ext.lstrip('.')}"
    with open(image_path, "rb") as f:
        data = base64.b64encode(f.read()).decode("utf-8")
    return {"type": "image_url", "image_url": {"url": f"data:{mime};base64,{data}"}}


def build_content(image_paths):
    prompt = "<PAGE><image>" * len(image_paths) + "Multi page parsing."
    return [encode_image(path) for path in image_paths] + [{"type": "text", "text": prompt}]


def generate(image_paths, crop_mode=True, tokens_per_page=8000):
    # A wider n-gram window for multi-page documents, as in our evaluation setup.
    window_size = 1024 if len(image_paths) > 1 else 128
    payload = {
        "model": "Unlimited-OCR-2",
        "messages": [{"role": "user", "content": build_content(image_paths)}],
        # Must be positive: it is the global output budget for the whole document.
        "max_tokens": tokens_per_page * len(image_paths),
        "temperature": 0,
        "skip_special_tokens": False,
        "images_config": {"crop_mode": crop_mode},  # True = gundam (tiles), False = base
        "custom_logit_processor": NGRAM_PROCESSOR,
        "custom_params": {
            "ngram_size": 35,
            "window_size": window_size,
            "page_token_id": 128827,   # <PAGE> marker id (= sparse_attention_config.page_token_id)
            "eos_token_id": 1,         # = eos_token_id in config.json
            "max_page_tokens": 20000,  # per-page output cap, <= 0 disables the page guard
        },
        "stream": True,
    }
    response = session.post(
        f"{server_url}/v1/chat/completions",
        headers={"Content-Type": "application/json"},
        data=json.dumps(payload),
        timeout=6000,
        stream=True,
    )
    response.raise_for_status()

    chunks = []
    for line in response.iter_lines(decode_unicode=True):
        if not line or not line.startswith("data:"):
            continue
        data = line[len("data:"):].strip()
        if data == "[DONE]":
            break
        event = json.loads(data)
        choices = event.get("choices") or []  # the final chunk carries usage only
        if not choices:
            continue
        delta = (choices[0].get("delta") or {}).get("content") or ""
        if delta:
            print(delta, end="", flush=True)
            chunks.append(delta)
    return "".join(chunks)


# Multi page: all pages of a document, parsed in one request
pages = [f"your_doc/page_{i}.jpg" for i in range(1, 8)]
generate(pages)

# Single page
generate(["your_doc/page_1.jpg"])

# PDF
generate(pdf_to_images("your_doc.pdf", dpi=300))

Evaluation

infer.py is the batch client we use for MPDocBench / MarathonDocBench: it sends each document to the SGLang server above as a single request and saves the raw output, which utils/process_unlimitedOCR2.py then converts to the evaluation Markdown.

python infer.py --url 127.0.0.1:10000 --benchmark MPDocBench \
    --dataset-root /path/to/MPDocBench --all --raw-dir raw --concurrency 4
python utils/process_unlimitedOCR2.py --infer-dir raw --out-dir pred

See utils/README.md for the output format, the merging rules and the post-processing details.

Visualization

Unlimited-horizon parsing in action: a full-length book parsed in a single request (replay).

Acknowledgement

We would like to thank Deepseek-OCR, Deepseek-OCR-2, PaddleOCR for their valuable models and ideas, and SGLang for the serving framework.

Citation

@misc{wei2026unlimitedocrworks2,
      title={Unlimited OCR Works 2}, 
      author={Baidu ERNIE Team},
      year={2026},
      eprint={2610.39308},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2610.39308}, 
}

@misc{yin2026unlimitedocrworks,
      title={Unlimited OCR Works}, 
      author={Youyang Yin and Huanhuan Liu and YY and Qunyi Xie and Chaorun Liu and Shiqi Yang and Shaohua Wang and Zhanlong Liu and Hao Zou and Jinyue Chen and Shu Wei and Jingjing Wu and Mingxin Huang and Zhen Wu and Guibin Wang and Tengyu Du and Lei Jia},
      year={2026},
      eprint={2606.23050},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2606.23050}, 
}