diff --git a/README.md b/README.md
index 3a537da..fdef6d4 100644
--- a/README.md
+++ b/README.md
@@ -6,16 +6,14 @@ tags:
 base_model:
 - Qwen/Qwen2-VL-2B-Instruct
 license: mit
-tasks:
-  - image-text-to-text
 ---
-[Github](https://github.com/showlab/ShowUI/tree/main) | [arXiv](https://arxiv.org/abs/2411.17465) | [HF Paper](https://huggingface.co/papers/2411.17465) | [Studio](https://www.modelscope.cn/studios/AI-ModelScope/ShowUI) | [Datasets](https://huggingface.co/datasets/showlab/ShowUI-desktop-8K) | [Quick Start](https://www.modelscope.cn/models/AI-ModelScope/ShowUI-2B) 
+[Github](https://github.com/showlab/ShowUI/tree/main) | [arXiv](https://arxiv.org/abs/2411.17465) | [HF Paper](https://huggingface.co/papers/2411.17465) | [Spaces](https://huggingface.co/spaces/showlab/ShowUI) | [Datasets](https://huggingface.co/datasets/showlab/ShowUI-desktop-8K) | [Quick Start](https://huggingface.co/showlab/ShowUI-2B) 
 <img src="examples/showui.jpg" alt="ShowUI" width="640">
 
 ShowUI is a lightweight (2B) vision-language-action model designed for GUI agents.
 
-##  Try our ModelScope Studio Demo
-https://www.modelscope.cn/studios/AI-ModelScope/ShowUI
+## 🤗 Try our HF Space Demo
+https://huggingface.co/spaces/showlab/ShowUI
 
 
 ## ⭐ Quick Start
@@ -49,7 +47,7 @@ model = Qwen2VLForConditionalGeneration.from_pretrained(
 min_pixels = 256*28*28
 max_pixels = 1344*28*28
 
-processor = AutoProcessor.from_pretrained("Qwen/Qwen2-VL-2B-Instruct", min_pixels=min_pixels, max_pixels=max_pixels)
+processor = AutoProcessor.from_pretrained("showlab/ShowUI-2B", min_pixels=min_pixels, max_pixels=max_pixels)
 ```
 
 2. **UI Grounding**
@@ -146,7 +144,7 @@ action_map = {
 ```python
 img_url = 'examples/chrome.png'
 split='web'
-system_prompt = _NAV_SYSTEM.format(_APP=split, _ACTION_SPACE=action_map[split])
+system_prompt = _NAV_SYSTEM.format(_APP=split, _ACTION_SPACE=action_map[split]) + _NAV_FORMAT
 query = "Search the weather for the New York city."
 
 messages = [