diff --git a/cpp/include/cudf/detail/utilities/cuda_memcpy.hpp b/cpp/include/cudf/detail/utilities/cuda_memcpy.hpp
index 3764a1f345b4..6c4997f5504b 100644
--- a/cpp/include/cudf/detail/utilities/cuda_memcpy.hpp
+++ b/cpp/include/cudf/detail/utilities/cuda_memcpy.hpp
@@ -1,5 +1,5 @@
 /*
- * SPDX-FileCopyrightText: Copyright (c) 2024-2026, NVIDIA CORPORATION.
+ * SPDX-FileCopyrightText: Copyright (c) 2024-2026, NVIDIA CORPORATION & AFFILIATES. All rights reserved.
  * SPDX-License-Identifier: Apache-2.0
  */
 
@@ -28,10 +28,8 @@ void cuda_memcpy_async_impl(
  * satisfied, but for host memory the caller must ensure the source is not freed before the stream
  * is synchronized.
  *
- * All copies share a single attribute entry (`cudaMemcpySrcAccessOrderStream` +
- * `cudaMemcpyFlagPreferOverlapWithCompute`). Per-copy attributes are not supported by this
- * wrapper; callers requiring different attributes per copy should call `cudaMemcpyBatchAsync`
- * directly.
+ * A batch uses `cudaMemcpyFlagPreferOverlapWithCompute` when every copy is 128 KiB or less. If
+ * any copy is larger, the batch uses `cudaMemcpyFlagDefault`.
  *
  * @param dsts Host pointer to a list of destination pointers.
  * @param srcs Host pointer to a list of source pointers.
diff --git a/cpp/src/utilities/cuda_memcpy.cu b/cpp/src/utilities/cuda_memcpy.cu
index 8e961a9cf633..542849eff92d 100644
--- a/cpp/src/utilities/cuda_memcpy.cu
+++ b/cpp/src/utilities/cuda_memcpy.cu
@@ -63,6 +63,8 @@ cudaError_t memcpy_batch_async(void* const* dsts,
 // cudaMemcpyBatchAsync does not support the default stream.
 #if CUDART_VERSION >= 13000
   if (!stream.is_default()) {
+    constexpr std::size_t prefer_overlap_threshold = 128 * 1024;
+
     // Filter out invalid copies (nullptr dst/src or size==0);
     // cudaMemcpyBatchAsync does not support these inputs
     auto is_invalid = [&](auto i) {
@@ -90,10 +92,14 @@ cudaError_t memcpy_batch_async(void* const* dsts,
       count = valid_dsts.size();
     }
 
-    cudaMemcpyAttributes attrs = {.srcAccessOrder = cudaMemcpySrcAccessOrderStream,
-                                  .flags          = cudaMemcpyFlagPreferOverlapWithCompute};
-    std::size_t attrs_idxs     = 0;
-    return cudaMemcpyBatchAsync(dsts, srcs, sizes, count, &attrs, &attrs_idxs, 1, stream.value());
+    unsigned int const flags =
+      std::ranges::any_of(std::ranges::views::iota(std::size_t{0}, count),
+                          [&](auto i) { return sizes[i] > prefer_overlap_threshold; })
+        ? cudaMemcpyFlagDefault
+        : cudaMemcpyFlagPreferOverlapWithCompute;
+    cudaMemcpyAttributes attrs = {.srcAccessOrder = cudaMemcpySrcAccessOrderStream, .flags = flags};
+    std::size_t attrs_idx      = 0;
+    return cudaMemcpyBatchAsync(dsts, srcs, sizes, count, &attrs, &attrs_idx, 1, stream.value());
   }
 #endif  // CUDART_VERSION >= 13000
   for (std::size_t i = 0; i < count; ++i) {