DeepSeek-OCR
 Collection
 2 items • Updated  •  34
- Libraries
-  Transformers How to use deepseek-ai/DeepSeek-OCR-2 with Transformers: # Use a pipeline as a high-level helper
from transformers import pipeline
pipe = pipeline("image-text-to-text", model="deepseek-ai/DeepSeek-OCR-2", trust_remote_code=True)# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("deepseek-ai/DeepSeek-OCR-2", trust_remote_code=True, device_map="auto")
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
-  vLLM  How to use deepseek-ai/DeepSeek-OCR-2 with vLLM: Install from pip and serve model# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "deepseek-ai/DeepSeek-OCR-2"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "deepseek-ai/DeepSeek-OCR-2",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'Use Dockerdocker model run hf.co/deepseek-ai/DeepSeek-OCR-2
-  SGLang  How to use deepseek-ai/DeepSeek-OCR-2 with SGLang: Install from pip and serve model# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "deepseek-ai/DeepSeek-OCR-2" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "deepseek-ai/DeepSeek-OCR-2",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'Use Docker imagesdocker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "deepseek-ai/DeepSeek-OCR-2" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "deepseek-ai/DeepSeek-OCR-2",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'
-  Docker Model Runner  How to use deepseek-ai/DeepSeek-OCR-2 with Docker Model Runner: docker model run hf.co/deepseek-ai/DeepSeek-OCR-2
🌟 Github | 📥 Model Download | 📄 Paper Link | 📄 Arxiv Paper Link |
DeepSeek-OCR 2: Visual Causal Flow
Explore more human-like visual encoding.
Inference using Huggingface transformers on NVIDIA GPUs. Requirements tested on python 3.12.9 + CUDA11.8：
torch==2.6.0
transformers==4.46.3
tokenizers==0.20.3
einops
addict
easydict
pip install flash-attn==2.7.3 --no-build-isolation
from transformers import AutoModel, AutoTokenizer
import torch
import os
os.environ["CUDA_VISIBLE_DEVICES"] = '0'
model_name = 'deepseek-ai/DeepSeek-OCR-2'
tokenizer = AutoTokenizer.from_pretrained(model_name, trust_remote_code=True)
model = AutoModel.from_pretrained(model_name, _attn_implementation='flash_attention_2', trust_remote_code=True, use_safetensors=True)
model = model.eval().cuda().to(torch.bfloat16)
# prompt = "<image>\nFree OCR. "
prompt = "<image>\n<|grounding|>Convert the document to markdown. "
image_file = 'your_image.jpg'
output_path = 'your/output/dir'
res = model.infer(tokenizer, prompt=prompt, image_file=image_file, output_path = output_path, base_size = 1024, image_size = 768, crop_mode=True, save_results = True)
Refer to 🌟GitHub for guidance on model inference acceleration and PDF processing, etc.
- Dynamic resolution
  - Default: (0-6)×768×768 + 1×1024×1024 — (0-6)×144 + 256 visual tokens ✅
# document: <image>\n<|grounding|>Convert the document to markdown.
# without layouts: <image>\nFree OCR.
We would like to thank DeepSeek-OCR, Vary, GOT-OCR2.0, MinerU, PaddleOCR for their valuable models and ideas.
We also appreciate the benchmark OmniDocBench.
@article{wei2025deepseek,
  title={DeepSeek-OCR: Contexts Optical Compression},
  author={Wei, Haoran and Sun, Yaofeng and Li, Yukun},
  journal={arXiv preprint arXiv:2510.18234},
  year={2025}
}
@article{wei2026deepseek,
  title={DeepSeek-OCR 2: Visual Causal Flow},
  author={Wei, Haoran and Sun, Yaofeng and Li, Yukun},
  journal={arXiv preprint arXiv:2601.20552},
  year={2026}
}
- Downloads last month
- 1,458,376
 Inference Providers NEW
  This model isn't deployed by any Inference Provider. 🙋 63 Ask for provider support
 Paper • 2601.20552 • Published  •  74
 Paper • 2510.18234 • Published  •  95
- llamaindex/ParseBench leaderboard
- Mean View evaluation resultssourcePipeline name: deepseekocr2_vllm41.2 *
- Text Content View evaluation resultssourcePipeline name: deepseekocr2_vllm82 *
- Text Formatting View evaluation resultssourcePipeline name: deepseekocr2_vllm54 *
- allenai/olmOCR-bench leaderboard
- Overall View evaluation resultssource76.3