docs: add verified vLLM Docker image to Deploy section

README.md

@@ -106,15 +106,11 @@ This model was obtained by quantizing the weights and activations of DeepSeek-V4
 
 ### Deploy with vLLM
 
-To deploy the quantized NVFP4 checkpoint with [vLLM](https://github.com/vllm-project/vllm), use the following command (you need 8xB200 GPUs):
+Verified serving this NVFP4 checkpoint on **GB300** with the `vllm/vllm-openai:nightly-aarch64` Docker image (vLLM 0.22.1rc1.dev504), tensor-parallel 4:
 
-```bash
-python -m vllm.entrypoints.cli.main serve \
-    nvidia/DeepSeek-V4-Flash-NVFP4 \
-    --tensor-parallel-size 8 \
-    --trust-remote-code \
-    --kv-cache-dtype fp8 \
-    --served-model-name nvfp4
+```sh
+vllm serve nvidia/DeepSeek-V4-Flash-NVFP4 \
+    --tensor-parallel-size 4 --trust-remote-code --kv-cache-dtype fp8
 ```
 
 ### Deploy with SGLang