diff --git a/tests/models/tapas/test_tokenization_tapas.py b/tests/models/tapas/test_tokenization_tapas.py
index d3af3276f751..b4cca18162d8 100644
--- a/tests/models/tapas/test_tokenization_tapas.py
+++ b/tests/models/tapas/test_tokenization_tapas.py
@@ -22,7 +22,7 @@
 import numpy as np
 import pandas as pd
 
-from transformers import AddedToken
+from transformers import AddedToken, is_torch_available
 from transformers.models.tapas.tokenization_tapas import (
     VOCAB_FILES_NAMES,
     BasicTokenizer,
@@ -44,6 +44,12 @@
 from ...test_tokenization_common import TokenizerTesterMixin, filter_non_english, merge_model_tokenizer_mappings
 
 
+if is_torch_available():
+    from transformers.pytorch_utils import is_torch_greater_or_equal_than_1_12
+else:
+    is_torch_greater_or_equal_than_1_12 = False
+
+
 @require_tokenizers
 @require_pandas
 class TapasTokenizationTest(TokenizerTesterMixin, unittest.TestCase):
@@ -1026,6 +1032,7 @@ def test_token_type_ids(self):
                 # Do the same test as modeling common.
                 self.assertIn(0, output["token_type_ids"][0])
 
+    @unittest.skipIf(not is_torch_greater_or_equal_than_1_12, reason="Tapas is only available in torch v1.12+")
     @require_torch
     @slow
     def test_torch_encode_plus_sent_to_model(self):