diff --git a/nds-h/README.md b/nds-h/README.md
index 76349b4..ef7b2ad 100644
--- a/nds-h/README.md
+++ b/nds-h/README.md
@@ -196,7 +196,7 @@ positional arguments:
 optional arguments:
   -h, --help            show this help message and exit
   --input_format {parquet,orc,avro,csv,json,iceberg,delta}
-                        type for input data source, e.g. parquet, orc, json, csv or iceberg, delta. Certain types are not fully supported by GPU reading, please refer to https://github.com/NVIDIA/spark-rapids/blob/branch-22.08/docs/compatibility.md for more details.
+                        type for input data source, e.g. parquet, orc, json, csv or iceberg, delta. Certain types are not fully supported by GPU reading, please refer to https://github.com/NVIDIA/spark-rapids/blob/branch-24.08/docs/compatibility.md for more details.
   --output_prefix OUTPUT_PREFIX
                         text to prepend to every output file (e.g., "hdfs:///ds-parquet")
   --output_format OUTPUT_FORMAT
diff --git a/nds-h/nds_h_power.py b/nds-h/nds_h_power.py
index eb55f72..be2dc98 100644
--- a/nds-h/nds_h_power.py
+++ b/nds-h/nds_h_power.py
@@ -284,7 +284,7 @@ def load_properties(filename):
     parser.add_argument('--input_format',
                         help='type for input data source, e.g. parquet, orc, json, csv or iceberg, delta. ' +
                         'Certain types are not fully supported by GPU reading, please refer to ' +
-                        'https://github.com/NVIDIA/spark-rapids/blob/branch-22.08/docs/compatibility.md ' +
+                        'https://github.com/NVIDIA/spark-rapids/blob/branch-24.08/docs/compatibility.md ' +
                         'for more details.',
                         choices=['parquet', 'orc', 'avro', 'csv', 'json', 'iceberg', 'delta'],
                         default='parquet')