diff --git a/benchmarks/benchmark_latency.py b/benchmarks/benchmark_latency.py
index e2d358ea..91510daf 100644
--- a/benchmarks/benchmark_latency.py
+++ b/benchmarks/benchmark_latency.py
@@ -169,8 +169,8 @@ if __name__ == '__main__':
         "--device",
         type=str,
         default="cuda",
-        choices=["cuda"],
-        help='device type for vLLM execution, supporting CUDA only currently.')
+        choices=["cuda", "cpu"],
+        help='device type for vLLM execution, supporting CUDA and CPU.')
     parser.add_argument('--block-size',
                         type=int,
                         default=16,
diff --git a/benchmarks/benchmark_throughput.py b/benchmarks/benchmark_throughput.py
index d6bf18c8..e7133827 100644
--- a/benchmarks/benchmark_throughput.py
+++ b/benchmarks/benchmark_throughput.py
@@ -329,8 +329,8 @@ if __name__ == "__main__":
         "--device",
         type=str,
         default="cuda",
-        choices=["cuda"],
-        help='device type for vLLM execution, supporting CUDA only currently.')
+        choices=["cuda", "cpu"],
+        help='device type for vLLM execution, supporting CUDA and CPU.')
     parser.add_argument(
         "--enable-prefix-caching",
         action='store_true',