diff --git a/ais_bench/benchmark/utils/tokenizer.py b/ais_bench/benchmark/utils/tokenizer.py
index e9cb76d65b2a21640eb20ba66c691568110d4b27..f039f5d8832a292217ec1394dccc23b65d1c2fe0 100644
--- a/ais_bench/benchmark/utils/tokenizer.py
+++ b/ais_bench/benchmark/utils/tokenizer.py
@@ -65,6 +65,9 @@ class MindformersTokenizer(Tokenizer):
     def decode(self, token_ids: list, skip_special_tokens=False) -> str:
         return self.tokenizer_model.decode(token_ids, skip_special_tokens=skip_special_tokens)
 
+    def batch_encode_plus(self, batch_text_or_text_pairs, *args, **kwargs):
+        return self.tokenizer_model.batch_encode_plus(batch_text_or_text_pairs, *args, **kwargs)
+
 
 class BenchmarkTokenizer:
     def __init__(self, model_path: str, tokenizer_type: str = None, trust_remote_code: bool = False, **kwargs):