diff --git a/python/sglang/srt/arg_groups/memory_hook.py b/python/sglang/srt/arg_groups/memory_hook.py
index 73f32960d043..87e4634c16ee 100644
--- a/python/sglang/srt/arg_groups/memory_hook.py
+++ b/python/sglang/srt/arg_groups/memory_hook.py
@@ -231,13 +231,11 @@ def handle_gpu_memory_settings(server_args: Any):
         decode_cuda_graph_config.max_bs = cfg.torch_compile_max_bs
 
     if prefill_cuda_graph_config.max_bs is None:
-        # Refer to pr #15927, by default we set the prefill max_bs to the chunked prefill size.
-        # For MLA backend, the introduction of piecewise cuda graph will influence the kernel dispatch difference compared to the original mode.
-        # To avoid the performance regression, we set max_bs to 2048 by default.
+        # Default to the chunked prefill size; cap MLA capture at 8192 tokens.
         if not use_mla_backend(server_args):
             prefill_cuda_graph_config.max_bs = cfg.chunked_prefill_size
         else:
-            prefill_cuda_graph_config.max_bs = 2048
+            prefill_cuda_graph_config.max_bs = min(8192, cfg.chunked_prefill_size)
 
         # For opt-in PP breakable graphs, capture small aggregate-token
         # buckets by default and leave larger forwards on the eager path.