diff --git a/ggml/src/ggml-webgpu/CMakeLists.txt b/ggml/src/ggml-webgpu/CMakeLists.txt
index 1503a1ef8ba6..d63c1da3ae89 100644
--- a/ggml/src/ggml-webgpu/CMakeLists.txt
+++ b/ggml/src/ggml-webgpu/CMakeLists.txt
@@ -81,3 +81,5 @@ endif()
 
 target_include_directories(ggml-webgpu PRIVATE ${SHADER_OUTPUT_DIR})
 target_link_libraries(ggml-webgpu PRIVATE ${DawnWebGPU_TARGET})
+
+target_compile_features(ggml-webgpu PRIVATE cxx_std_20)
diff --git a/ggml/src/ggml-webgpu/ggml-webgpu.cpp b/ggml/src/ggml-webgpu/ggml-webgpu.cpp
index 377f3d702c02..0f76b87852f2 100644
--- a/ggml/src/ggml-webgpu/ggml-webgpu.cpp
+++ b/ggml/src/ggml-webgpu/ggml-webgpu.cpp
@@ -1603,7 +1603,7 @@ static webgpu_encoded_op ggml_webgpu_mul_mat(webgpu_context & ctx,
                                              ggml_tensor *    src1,
                                              ggml_tensor *    dst) {
     // Determine if this is a mat-vec operation
-    bool use_mat_vec = (dst->ne[1] <= 4);
+    bool use_mat_vec = (dst->ne[1] <= 8);
 
     // use MMVQ path for mat-vec
     bool use_mmvq = ggml_webgpu_can_use_mmvq(src0, src1, ctx->global_ctx->capabilities.supports_dot_product,