diff --git a/python/llm/dev/benchmark/all-in-one/run.py b/python/llm/dev/benchmark/all-in-one/run.py
index 993a4fe0..05574365 100644
--- a/python/llm/dev/benchmark/all-in-one/run.py
+++ b/python/llm/dev/benchmark/all-in-one/run.py
@@ -96,6 +96,7 @@ def run_model(repo_id, test_api, in_out_pairs, local_model_hub=None, warm_up=1,
                             round(np.mean(result[in_out_pair], axis=0)[1]*1000.0, 2),
                             round(np.mean(result[in_out_pair], axis=0)[2]*1000.0, 2),
                             in_out_pair,
+                            batch_size,
                             f'{int(np.mean(result[in_out_pair], axis=0)[3])}' +
                             f'-{int(np.mean(result[in_out_pair], axis=0)[4])}',
                             num_beams,
@@ -445,8 +446,8 @@ def run_transformer_int4_gpu(repo_id,
                     csv_writer = csv.writer(file)
                     file.seek(0, os.SEEK_END)
                     if file.tell() == 0:
-                        csv_writer.writerow(["","model","1st token avg latency (ms)","2+ avg latency (ms/token)","encoder time (ms)","input/output tokens","actual input/output tokens","num_beams","low_bit","cpu_embedding","peak mem (GB)"])
-                    csv_writer.writerow(['', repo_id, first_token_latency, rest_token_latency, encoder_time, input_output_tokens, actual_input_output_tokens, num_beams, low_bit, '', peak_mem])
+                        csv_writer.writerow(["","model","1st token avg latency (ms)","2+ avg latency (ms/token)","encoder time (ms)","input/output tokens", "batch_size", "actual input/output tokens","num_beams","low_bit","cpu_embedding","peak mem (GB)"])
+                    csv_writer.writerow(['', repo_id, first_token_latency, rest_token_latency, encoder_time, input_output_tokens, batch_size, actual_input_output_tokens, num_beams, low_bit, '', peak_mem])
 
     model.to('cpu')
     torch.xpu.synchronize()
@@ -960,7 +961,7 @@ if __name__ == '__main__':
             run_model(model, api, in_out_pairs, conf['local_model_hub'], conf['warm_up'], conf['num_trials'], conf['num_beams'],
                       conf['low_bit'], conf['cpu_embedding'], conf['batch_size'])
         df = pd.DataFrame(results, columns=['model', '1st token avg latency (ms)', '2+ avg latency (ms/token)', 'encoder time (ms)',
-                                            'input/output tokens', 'actual input/output tokens', 'num_beams', 'low_bit', 'cpu_embedding',
+                                            'input/output tokens', 'batch_size', 'actual input/output tokens', 'num_beams', 'low_bit', 'cpu_embedding',
                                             'peak mem (GB)'])
         df.to_csv(csv_name)
         results = []