From 7dde3438e204d14b351b6e9ae2a8925392b5b798 Mon Sep 17 00:00:00 2001
From: Yi Zhong <207368749+vincentzed@users.noreply.github.com>
Date: Thu, 15 Jan 2026 15:25:20 -0500
Subject: [PATCH] Show how to use cu13 image with B300 (#17170)

Signed-off-by: vincentzed <207368749+vincentzed@users.noreply.github.com>
---
 docs/get_started/install.md | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/docs/get_started/install.md b/docs/get_started/install.md
index 509a34248..733cfb250 100644
--- a/docs/get_started/install.md
+++ b/docs/get_started/install.md
@@ -21,6 +21,10 @@ For example, on GB200, you will need to do the following. Otherwise, it will ins
 uv pip install "sglang" --extra-index-url https://download.pytorch.org/whl/cu129
 ```
 
+On B300 (SM103), we recommend using the image at lmsysorg/sglang:dev-cu13.
+Please, do not re-install the project as editable inside the docker image, since it will override the version of
+libraries specified by the cu13 docker image.
+
 **Quick fixes to common problems**
 
 - If you encounter `OSError: CUDA_HOME environment variable is not set`. Please set it to your CUDA install root with either of the following solutions: