Update README.md

README.md

@@ -255,29 +255,49 @@ pip install modelscope
 python download_model.py -t modelscope -n MonkeyOCRv2-B-Parsing # or MonkeyOCRv2-S-Parsing
 ```
 
-#### 3. Inference
+#### 3. vLLM Serving
+You should start a vLLM service before parsing documents:
+```bash
+cd parsing
+python serve.py -m ../model_weight/MonkeyOCRv2-B-Parsing -p 8888
+# Show help messages
+python serve.py -h
+```
+
+#### 4. Inference
+You can parse documents using CLI or serve with demo and FastAPI.
+
+##### 4.1 Parse using CLI
 Parse a single document or a directory containing PDFs or images:
 ```bash
 cd parsing
 python parse.py \
-    -i ../images_test/ar.JPEG \
+    -i ../images_test \
     -o output/test \
-    -m ../model_weight/MonkeyOCRv2-B-Parsing \
-    -g 500 \
+    -s http://127.0.0.1:8888 \
     --draw-layout \
     --skip-processed
 # Show help messages
 python parse.py -h
 ```
 
-#### 4. Web Demo
-Start gradio web demo:
+##### 4.2 Serve with Web Demo
 ```bash
 cd parsing
-python demo/gradio_demo.py \
-    --model-path ../model_weight/MonkeyOCRv2-B-Parsing \
-    --output-dir output/demo_outputs
+python demo/gradio_demo.py -s http://127.0.0.1:8888 -p 8891
+# Show help messages
+python demo/gradio_demo.py -h
 ```
+You can access the web demo at http://localhost:8891.
+
+##### 4.3 Serve with FastAPI
+```bash
+cd parsing
+python fastapi/main.py -s http://127.0.0.1:8888 -p 8000
+# Show help messages
+python fastapi/main.py -h
+```
+You can access the API documentation at http://localhost:8000/docs to explore available endpoints.
 
 ### Document Understanding
 #### 1. Install
@@ -997,12 +1017,12 @@ Our model supports robust document parsing in real-world scenarios across 17 lan
 <tr>
 <td>dots.mocr</td><td>3B</td><td>1.2B</td><td>1.8B</td><td>80.5</td><td><strong>90.5</strong></td><td>77.2</td><td>81.7</td><td>82.6</td><td><strong>87.4</strong></td><td>71.3</td><td>70.1</td><td>84.5</td><td><strong>89.3</strong></td><td>83.2</td><td>86.8</td><td>79.9</td><td>79.2</td><td>83.3</td><td>83.6</td><td><strong>75.0</strong></td><td>78.7</td><td>71.2</td><td>77.9</td><td><u>84.6</u></td><td><strong>79.6</strong></td>
 </tr>
-    <tr>
-      <td><strong>MonkeyOCRv2-S-Parsing<a href="https://huggingface.co/zenosai/MonkeyOCRv2-S-Parsing">🤗</a></strong></td><td>0.6B</td><td>0.03B</td><td>0.6B</td><td><u>82.5</u></td><td>87.9</td><td><u>80.7</u></td><td><u>83.2</u></td><td><u>87.3</u></td><td>83.6</td><td><strong>76.8</strong></td><td>73.6</td><td><u>85.4</u></td><td>87.2</td><td><u>85.5</u></td><td>87.4</td><td>81.9</td><td><u>81.7</u></td><td><strong>91.2</strong></td><td><u>87.1</u></td><td>69.9</td><td><strong>88.7</strong></td><td><u>78.0</u></td><td><u>79.8</u></td><td>84.4</td><td>74.7</td>
-    </tr>
-    <tr>
-      <td><strong>MonkeyOCRv2-B-Parsing<a href="https://huggingface.co/zenosai/MonkeyOCRv2-B-Parsing">🤗</a><strong></td><td>0.7B</td><td>0.1B</td><td>0.6B</td><td><strong>83.3</strong></td><td><u>88.1</u></td><td><strong>81.7</strong></td><td><strong>84.2</strong></td><td><strong>87.7</strong></td><td>84.5</td><td>75.2</td><td><strong>78.4</strong></td><td><strong>86.5</strong></td><td><u>88.6</u></td><td><strong>86.1</strong></td><td>87.9</td><td>83.2</td><td><strong>82.1</strong></td><td><u>90.7</u></td><td><strong>87.2</strong></td><td>71.9</td><td><u>87.6</u></td><td><strong>80.1</strong></td><td><strong>80.8</strong></td><td>83.6</td><td>75.3</td>
-    </tr>
+<tr>
+<td><strong>MonkeyOCRv2-S-Parsing<a href="https://huggingface.co/zenosai/MonkeyOCRv2-S-Parsing">🤗</a></strong></td><td>0.6B</td><td>0.03B</td><td>0.6B</td><td><u>82.5</u></td><td>87.9</td><td><u>80.7</u></td><td><u>83.2</u></td><td><u>87.3</u></td><td>83.6</td><td><strong>76.8</strong></td><td>73.6</td><td><u>85.4</u></td><td>87.2</td><td><u>85.5</u></td><td>87.4</td><td>81.9</td><td><u>81.7</u></td><td><strong>91.2</strong></td><td><u>87.1</u></td><td>69.9</td><td><strong>88.7</strong></td><td><u>78.0</u></td><td><u>79.8</u></td><td>84.4</td><td>74.7</td>
+</tr>
+<tr>
+<td><strong>MonkeyOCRv2-B-Parsing<a href="https://huggingface.co/zenosai/MonkeyOCRv2-B-Parsing">🤗</a><strong></td><td>0.7B</td><td>0.1B</td><td>0.6B</td><td><strong>83.3</strong></td><td><u>88.1</u></td><td><strong>81.7</strong></td><td><strong>84.2</strong></td><td><strong>87.7</strong></td><td>84.5</td><td>75.2</td><td><strong>78.4</strong></td><td><strong>86.5</strong></td><td><u>88.6</u></td><td><strong>86.1</strong></td><td>87.9</td><td>83.2</td><td><strong>82.1</strong></td><td><u>90.7</u></td><td><strong>87.2</strong></td><td>71.9</td><td><u>87.6</u></td><td><strong>80.1</strong></td><td><strong>80.8</strong></td><td>83.6</td><td>75.3</td>
+</tr>
   </tbody>
 </table>
 
@@ -1191,16 +1211,4 @@ This multi-expert agreement strategy reduces model-specific annotation errors an
 - **Qwen3**: https://github.com/QwenLM/Qwen3
 
 ## Copyright
-We warmly welcome your feedback, suggestions, and contributions, which are essential to the continued development and improvement of our framework. Note: This model is intended for academic research and non-commercial use only. For any questions, please contact us at xbai@hust.edu.cn or ylliu@hust.edu.cn.
-
-## Citation
-If you find **MonkeyOCRv2** useful, consider citing our work: 
-
-```bibtex
-@article{liu2026monkeyocrv2,
-  title={MonkeyOCRv2: A Visual-Text Foundation Model for Document AI},
-  author={Yuliang Liu and Zhang Li and Ziyang Zhang and Shuo Zhang and Qiang Liu and Jiajun Song and Zidun Guo and Xinhan Wang and Handong Zheng and Yang Liu and Dongliang Luo and Zhiyin Ma and Jiarui Zhang and Xiang Bai},
-  journal={arXiv preprint arXiv:2607.11562},
-  year={2026}
-}
-```
+We warmly welcome your feedback, suggestions, and contributions, which are essential to the continued development and improvement of our framework. Note: This model is intended for academic research and non-commercial use only. For any questions, please contact us at xbai@hust.edu.cn or ylliu@hust.edu.cn.