diff --git a/cpp/tensorrt_llm/kernels/trtllmGenKernels/fmha/fmhaKernels.h b/cpp/tensorrt_llm/kernels/trtllmGenKernels/fmha/fmhaKernels.h
index 423f3de064cf..3429c8819654 100644
--- a/cpp/tensorrt_llm/kernels/trtllmGenKernels/fmha/fmhaKernels.h
+++ b/cpp/tensorrt_llm/kernels/trtllmGenKernels/fmha/fmhaKernels.h
@@ -306,7 +306,7 @@ class TllmGenFmhaKernel
     inline static std::vector<int> const kDefaultWarmupPrefillBatchSizeCandidates
         = {1, 2, 3, 4, 6, 8, 12, 16, 24, 32, 48, 64, 128, 256};
     inline static std::vector<int> const kDefaultWarmupSeqLenQkvCandidates
-        = {1, 128, 512, 1024, 2048, 4096, 8192, 16384, 32768};
+        = {1, 128, 256, 512, 1024, 2048, 3072, 4096, 8192, 16384, 32768};
 
     static std::vector<int> makeWarmupCandidateSizes(std::vector<int> const& defaultCandidateSizes, int maxSize)
     {