diff --git a/assets/llavabench_compare_3.png b/assets/llavabench_compare_3.png
index 2b45480..1423333 100644
Binary files a/assets/llavabench_compare_3.png and b/assets/llavabench_compare_3.png differ
diff --git a/assets/llavabench_compare_phi3.png b/assets/llavabench_compare_phi3.png
new file mode 100644
index 0000000..b514059
Binary files /dev/null and b/assets/llavabench_compare_phi3.png differ
diff --git a/docs/compare_with_phi-3_vision.md b/docs/compare_with_phi-3_vision.md
index 3df45e2..aeff647 100644
--- a/docs/compare_with_phi-3_vision.md
+++ b/docs/compare_with_phi-3_vision.md
@@ -41,7 +41,7 @@ MiniCPM-Llama3-V 2.5 exhibits **stronger multilingual** capabilities compared wi
 MiniCPM-Llama3-V 2.5 在对话和推理评测榜单 LLaVA Bench 上展现出了比 Phi-3-vision-128K-Instruct **更强的多语言的性能**。
 
 <div align="center">
-    <img src="../assets/llavabench_compare_3.png" width="85%" />
+    <img src="../assets/llavabench_compare_phi-3.png" width="85%" />
     <br>
     Evaluation results of LLaVABench in multiple languages
     <br>