diff --git a/server/deps/llama.cpp/ggml/src/ggml-cuda/ggml-cuda.cu b/server/deps/llama.cpp/ggml/src/ggml-cuda/ggml-cuda.cu
index f8fc6aa2a..a1c178fde 100644
--- a/server/deps/llama.cpp/ggml/src/ggml-cuda/ggml-cuda.cu
+++ b/server/deps/llama.cpp/ggml/src/ggml-cuda/ggml-cuda.cu
@@ -2816,7 +2816,19 @@ static bool ggml_cuda_try_fuse_mul_mat_glu(
         }
 
         const int64_t ncols = ids ? src1->ne[2] : src1->ne[1];
-        if (ggml_cuda_should_use_mmq(
+#ifdef GGML_CUDA_FORCE_CUBLAS
+        // The global force-cuBLAS policy is authoritative over this RDNA 3.5 default.
+        constexpr bool default_ds4_mix_gate_up_mmq = false;
+#else
+        const char * mix_mmq = std::getenv("DFLASH_DS4_MIX_MMQ_PREFILL");
+        const bool default_ds4_mix_gate_up_mmq =
+            src0->type == GGML_TYPE_Q2_1_ROCMFP2_MIX &&
+            ids != nullptr &&
+            GGML_CUDA_CC_IS_RDNA3_5(cc) &&
+            (!mix_mmq || !(mix_mmq[0] == '0' && mix_mmq[1] == '\0'));
+#endif
+        if (default_ds4_mix_gate_up_mmq ||
+            ggml_cuda_should_use_mmq(
                 src0->type, cc, ncols,
                 ids ? src0->ne[2] : /*n_experts=*/0)) {
             ggml_cuda_mul_mat_q_pair(