diff --git a/docs/en/integrations/tensorrt.md b/docs/en/integrations/tensorrt.md
index d7448c2d..583bef10 100644
--- a/docs/en/integrations/tensorrt.md
+++ b/docs/en/integrations/tensorrt.md
@@ -13,7 +13,7 @@ By using the TensorRT export format, you can enhance your [Ultralytics YOLOv8](h
 ## TensorRT
 
 <p align="center">
-  <img width="100%" src="https://docs.nvidia.com/deeplearning/tensorrt/archives/tensorrt-601/tensorrt-developer-guide/graphics/whatistrt2.png" alt="TensorRT Overview">
+  <img width="100%" src="https://github.com/ultralytics/yolov5/releases/download/v1.0/tensorrt-overview.jpg" alt="TensorRT Overview">
 </p>
 
 [TensorRT](https://developer.nvidia.com/tensorrt), developed by NVIDIA, is an advanced software development kit (SDK) designed for high-speed deep learning inference. It's well-suited for real-time applications like object detection.