diff --git a/be/src/common/config.cpp b/be/src/common/config.cpp
index 75495dcb1a2106..474f74433b501e 100644
--- a/be/src/common/config.cpp
+++ b/be/src/common/config.cpp
@@ -524,10 +524,10 @@ DEFINE_mDouble(compaction_promotion_ratio, "0.05");
 DEFINE_mInt64(compaction_promotion_min_size_mbytes, "128");
 
 // When output rowset of cumulative compaction total version count (end_version - start_version)
-// exceed this config count, the rowset will be moved to base compaction
-// NOTE: this config will work for unique key merge-on-write table only, to reduce version count
-// related cost on delete bitmap more effectively.
-DEFINE_mInt64(compaction_promotion_version_count, "1000");
+// exceed this config count, the rowset will be moved to base compaction.
+// NOTE: this config only works for unique key merge-on-write tables. The default maximum value
+// disables version-count-based promotion while retaining the option to enable it when needed.
+DEFINE_mInt64(compaction_promotion_version_count, "9223372036854775807");
 
 // The lower bound size to do cumulative compaction. When total disk size of candidate rowsets is less than
 // this size, size_based policy may not do to cumulative compaction. The unit is m byte.
diff --git a/be/src/common/config.h b/be/src/common/config.h
index b9e33288396033..1c93d4a4f39ee3 100644
--- a/be/src/common/config.h
+++ b/be/src/common/config.h
@@ -614,9 +614,9 @@ DECLARE_mDouble(compaction_promotion_ratio);
 DECLARE_mInt64(compaction_promotion_min_size_mbytes);
 
 // When output rowset of cumulative compaction total version count (end_version - start_version)
-// exceed this config count, the rowset will be moved to base compaction
-// NOTE: this config will work for unique key merge-on-write table only, to reduce version count
-// related cost on delete bitmap more effectively.
+// exceed this config count, the rowset will be moved to base compaction.
+// NOTE: this config only works for unique key merge-on-write tables. The default maximum value
+// disables version-count-based promotion while retaining the option to enable it when needed.
 DECLARE_mInt64(compaction_promotion_version_count);
 
 // The lower bound size to do cumulative compaction. When total disk size of candidate rowsets is less than
diff --git a/be/test/cloud/cloud_cumulative_compaction_policy_test.cpp b/be/test/cloud/cloud_cumulative_compaction_policy_test.cpp
index 35a2598617fca9..0bb4dbfc3bf63d 100644
--- a/be/test/cloud/cloud_cumulative_compaction_policy_test.cpp
+++ b/be/test/cloud/cloud_cumulative_compaction_policy_test.cpp
@@ -212,6 +212,19 @@ TEST_F(TestCloudSizeBasedCumulativeCompactionPolicy, new_cumulative_point) {
     EXPECT_EQ(policy.new_cumulative_point(&_tablet, output_rowset, version, 2), 6);
 }
 
+TEST_F(TestCloudSizeBasedCumulativeCompactionPolicy,
+       new_cumulative_point_does_not_promote_small_mow_rowset_with_default_version_limit) {
+    CloudTablet tablet(_engine, _tablet_meta);
+    tablet._tablet_meta->_enable_unique_key_merge_on_write = true;
+    tablet._base_size = kGiB;
+
+    CloudSizeBasedCumulativeCompactionPolicy policy;
+    RowsetSharedPtr output_rowset = create_rowset(Version(2, 1003), 1, false, kMiB);
+    Version last_delete_version {-1, -1};
+
+    EXPECT_EQ(2, policy.new_cumulative_point(&tablet, output_rowset, last_delete_version, 2));
+}
+
 TEST_F(TestCloudSizeBasedCumulativeCompactionPolicy,
        pick_input_rowsets_notready_keeps_latest_versions) {
     auto base_rowset = create_rowset(Version(0, 1), 1, false, kGiB);
diff --git a/be/test/storage/compaction/cumulative_compaction_policy_test.cpp b/be/test/storage/compaction/cumulative_compaction_policy_test.cpp
index ad1c536451da7f..a82c52719b1b86 100644
--- a/be/test/storage/compaction/cumulative_compaction_policy_test.cpp
+++ b/be/test/storage/compaction/cumulative_compaction_policy_test.cpp
@@ -1238,6 +1238,29 @@ TEST_F(TestSizeBasedCumulativeCompactionPolicy,
     EXPECT_EQ(0, compaction_score);
 }
 
+TEST_F(TestSizeBasedCumulativeCompactionPolicy,
+       update_cumulative_point_does_not_promote_small_mow_rowset_with_default_version_limit) {
+    _tablet_meta->set_enable_unique_key_merge_on_write(true);
+    TabletSharedPtr tablet(
+            new Tablet(_engine, _tablet_meta, nullptr, CUMULATIVE_SIZE_BASED_POLICY));
+    ASSERT_TRUE(tablet->init().ok());
+    tablet->set_cumulative_layer_point(1);
+    tablet->set_cumulative_promotion_size(64L * kMiB);
+
+    RowsetMetaSharedPtr output_meta(new RowsetMeta());
+    init_rs_meta(output_meta, 1, 1002);
+    output_meta->set_total_disk_size(kMiB);
+    RowsetSharedPtr output_rowset;
+    ASSERT_TRUE(RowsetFactory::create_rowset(nullptr, "", output_meta, &output_rowset).ok());
+
+    std::vector<RowsetSharedPtr> input_rowsets;
+    Version last_delete_version {-1, -1};
+    tablet->_cumulative_compaction_policy->update_cumulative_point(
+            tablet.get(), input_rowsets, output_rowset, last_delete_version);
+
+    EXPECT_EQ(1, tablet->cumulative_layer_point());
+}
+
 TEST_F(TestSizeBasedCumulativeCompactionPolicy,
        pick_input_rowsets_large_head_single_overlapping_tail_selected) {
     std::vector<RowsetMetaSharedPtr> rs_metas;