diff --git a/mteb/benchmarks/_create_table.py b/mteb/benchmarks/_create_table.py
index 23e4296339..5acaf039e7 100644
--- a/mteb/benchmarks/_create_table.py
+++ b/mteb/benchmarks/_create_table.py
@@ -303,6 +303,7 @@ def _create_per_language_table_from_benchmark_results(
 
 def _create_summary_table_mean_public_private(
     benchmark_results: BenchmarkResults,
+    exclude_private_from_borda: bool = False,
 ) -> pd.DataFrame:
     """Create summary table from BenchmarkResults.
 
@@ -311,6 +312,7 @@ def _create_summary_table_mean_public_private(
 
     Args:
         benchmark_results: BenchmarkResults object containing model results
+        exclude_private_from_borda: If True, calculate Borda rank using only public tasks
 
     Returns:
         DataFrame with model summaries, ready for styling in the leaderboard
@@ -356,7 +358,11 @@ def _create_summary_table_mean_public_private(
     joint_table = joint_table.drop(models_to_remove, axis=0)
     joint_table.insert(0, "mean(public)", public_mean)
     joint_table.insert(1, "mean(private)", private_mean)
-    joint_table["borda_rank"] = _get_borda_rank(per_task)
+    if exclude_private_from_borda:
+        borda_per_task = per_task[public_task_name]
+    else:
+        borda_per_task = per_task
+    joint_table["borda_rank"] = _get_borda_rank(borda_per_task)
     joint_table = joint_table.sort_values("borda_rank", ascending=True)
     joint_table = joint_table.reset_index()
 
diff --git a/mteb/benchmarks/benchmark.py b/mteb/benchmarks/benchmark.py
index 3b6dd4aa0e..74c93f74b8 100644
--- a/mteb/benchmarks/benchmark.py
+++ b/mteb/benchmarks/benchmark.py
@@ -123,7 +123,9 @@ def _create_summary_table(
             _create_summary_table_mean_public_private,
         )
 
-        joint_table = _create_summary_table_mean_public_private(benchmark_results)
+        joint_table = _create_summary_table_mean_public_private(
+            benchmark_results, exclude_private_from_borda=True
+        )
         # issue 3902: temporary remove the private column from RTEB summary table
         if "Mean (Private)" in joint_table.columns:
             joint_table = joint_table.drop(columns=["Mean (Private)"])