diff --git a/README.md b/README.md
index ecd3bf4..6dc106a 100644
--- a/README.md
+++ b/README.md
@@ -1,12 +1,11 @@
 <div align="center">
 <h1>VGGT: Visual Geometry Grounded Transformer</h1>
 
-<!-- <a href=""><img src='https://img.shields.io/badge/arXiv-VGGT' alt='Paper PDF'></a> -->
-<!-- <a href=''><img src='https://img.shields.io/badge/Project_Page-green' alt='Project Page'></a> -->
-
 <a href="https://jytime.github.io/data/VGGT_CVPR25.pdf" target="_blank" rel="noopener noreferrer">
   <img src="https://img.shields.io/badge/Paper-VGGT" alt="Paper PDF">
 </a>
+<a href="https://arxiv.org/abs/2503.11651"><img src="https://img.shields.io/badge/arXiv-2503.11651-b31b1b" alt="arXiv"></a>
+<a href="https://vgg-t.github.io/"><img src="https://img.shields.io/badge/Project_Page-green" alt="Project Page"></a>
 <a href='https://huggingface.co/spaces/facebook/vggt'><img src='https://img.shields.io/badge/%F0%9F%A4%97%20Hugging%20Face-Demo-blue'></a>
 
 
@@ -24,7 +23,6 @@
   year={2025}
 }
 ```
-
 ## Overview
 
 Visual Geometry Grounded Transformer (VGGT, CVPR 2025) is a feed-forward neural network that directly infers all key 3D attributes of a scene, including extrinsic and intrinsic camera parameters, point maps, depth maps, and 3D point tracks, **from one, a few, or hundreds of its views, within seconds**.