diff --git a/python/llm/test/inference_gpu/test_optimize_model.py b/python/llm/test/inference_gpu/test_optimize_model.py
new file mode 100644
index 00000000..515686f7
--- /dev/null
+++ b/python/llm/test/inference_gpu/test_optimize_model.py
@@ -0,0 +1,58 @@
+#
+# Copyright 2016 The BigDL Authors.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+#
+
+import pytest
+import os
+
+from bigdl.llm.transformers import AutoModelForCausalLM, AutoModel
+from transformers import LlamaTokenizer, AutoTokenizer
+
+device = os.environ['DEVICE']
+print(f'Running on {device}')
+if device == 'xpu':
+    import intel_extension_for_pytorch as ipex
+
+prompt = "Once upon a time, there existed a little girl who liked to have adventures. She wanted to go to places and meet new people, and have fun"
+
+@pytest.mark.parametrize('Model, Tokenizer, model_path',[
+    (AutoModelForCausalLM, AutoTokenizer, os.environ.get('MPT_7B_ORIGIN_PATH')),
+    (AutoModelForCausalLM, AutoTokenizer, os.environ.get('FALCON_7B_ORIGIN_PATH')),
+    ])
+def test_optimize_model(Model, Tokenizer, model_path):
+    tokenizer = Tokenizer.from_pretrained(model_path, trust_remote_code=True)
+    input_ids = tokenizer.encode(prompt, return_tensors="pt").to(device)
+
+    model = Model.from_pretrained(model_path,
+                                load_in_4bit=True,
+                                optimize_model=False,
+                                trust_remote_code=True)
+    model = model.to(device)
+    logits_base_model = (model(input_ids)).logits
+
+    model = Model.from_pretrained(model_path,
+                                load_in_4bit=True,
+                                optimize_model=True,
+                                trust_remote_code=True)
+    model = model.to(device)
+    logits_optimized_model = (model(input_ids)).logits
+    
+    diff = abs(logits_base_model - logits_optimized_model).flatten()
+
+    assert any(diff) is False
+
+
+if __name__ == '__main__':
+    pytest.main([__file__])
diff --git a/python/llm/test/inference_gpu/test_transformers_api.py b/python/llm/test/inference_gpu/test_transformers_api.py
index 69f2578d..6ac9591a 100644
--- a/python/llm/test/inference_gpu/test_transformers_api.py
+++ b/python/llm/test/inference_gpu/test_transformers_api.py
@@ -27,12 +27,13 @@ if device == 'xpu':
     import intel_extension_for_pytorch as ipex
 
 @pytest.mark.parametrize('prompt, answer', [
-    ('What is the capital of France?\n\n','Paris')
+    ('What is the capital of France?\n\n', 'Paris')
     ])
 @pytest.mark.parametrize('Model, Tokenizer, model_path',[
     (AutoModelForCausalLM, LlamaTokenizer, os.environ.get('LLAMA2_7B_ORIGIN_PATH')),
     (AutoModel, AutoTokenizer, os.environ.get('CHATGLM2_6B_ORIGIN_PATH')),
     (AutoModelForCausalLM, AutoTokenizer, os.environ.get('FALCON_7B_ORIGIN_PATH')),
+    (AutoModelForCausalLM, AutoTokenizer, os.environ.get('MPT_7B_ORIGIN_PATH')),
     ])
 def test_completion(Model, Tokenizer, model_path, prompt, answer):
     tokenizer = Tokenizer.from_pretrained(model_path, trust_remote_code=True)