diff --git a/csrc/kernels/legacy/internode_ll.cu b/csrc/kernels/legacy/internode_ll.cu
index b6be356a..cc36718a 100644
--- a/csrc/kernels/legacy/internode_ll.cu
+++ b/csrc/kernels/legacy/internode_ll.cu
@@ -702,7 +702,7 @@ void dispatch(void* packed_recv_x,
               int num_device_sms,
               cudaStream_t stream,
               int phases) {
-    constexpr int kNumMaxTopK = 11;
+    constexpr int kNumMaxTopK = 16;  // Kimi-K3 routes top-16
     const int num_warp_groups = ceil_div(num_experts, num_device_sms);
     const int num_warps_per_group = 32 / num_warp_groups;
     EP_HOST_ASSERT(num_warp_groups > 0 and num_warps_per_group > 0);
@@ -1880,7 +1880,7 @@ void combine(void* combined_x,
         );
     }
 
-    constexpr int kNumMaxTopk = 11;
+    constexpr int kNumMaxTopk = 16;  // Kimi-K3 routes top-16
     const int num_warp_groups = ceil_div(num_experts, num_device_sms);
     const int num_warps_per_group = 32 / num_warp_groups;
     const int num_recv_per_sm = ceil_div(num_combined_tokens, num_device_sms);
diff --git a/csrc/kernels/legacy/launch.cuh b/csrc/kernels/legacy/launch.cuh
index 60e30803..36351e52 100644
--- a/csrc/kernels/legacy/launch.cuh
+++ b/csrc/kernels/legacy/launch.cuh
@@ -118,6 +118,8 @@
             case_macro(2560);                               \
         case 3072:                                          \
             case_macro(3072); /* for gpt-oss */             \
+        case 3584:                                          \
+            case_macro(3584); /* for kimi-k3 */             \
         case 4096:                                          \
             case_macro(4096);                               \
         case 5120:                                          \