diff --git a/demo-serving/op/bert_service_op.cpp b/demo-serving/op/bert_service_op.cpp
index e4618dc7de989536397af4135f9c2e5f6e5164a2..f4713782bf54e894efdf391c05231d77accf7019 100644
--- a/demo-serving/op/bert_service_op.cpp
+++ b/demo-serving/op/bert_service_op.cpp
@@ -31,7 +31,7 @@ using baidu::paddle_serving::predictor::bert_service::Embedding_values;
 extern int64_t MAX_SEQ_LEN = 128;
 const bool POOLING = true;
 const int LAYER_NUM = 12;
-const int EMB_SIZE = 768;
+extern int EMB_SIZE = 768;
 
 int BertServiceOp::inference() {
   timeval op_start;
@@ -49,6 +49,7 @@ int BertServiceOp::inference() {
   }
 
   MAX_SEQ_LEN = req->instances(0).max_seq_len();
+  EMB_SIZE = req->instances(0).emb_size();
 
   paddle::PaddleTensor src_ids;
   paddle::PaddleTensor pos_ids;
diff --git a/demo-serving/proto/bert_service.proto b/demo-serving/proto/bert_service.proto
index ce1ceeb5ee440c5fd6c8ff2e573d46c30fb4e8ff..3eba501317c2533ad8c02d79d5eca2dc6682a65e 100644
--- a/demo-serving/proto/bert_service.proto
+++ b/demo-serving/proto/bert_service.proto
@@ -26,6 +26,7 @@ message BertReqInstance {
   repeated int64 position_ids = 3;
   repeated float input_masks = 4;
   required int64 max_seq_len = 5;
+  required int64 emb_size = 6;
 };
 
 message Request { repeated BertReqInstance instances = 1; };
@@ -34,7 +35,11 @@ message Embedding_values { repeated float values = 1; };
 
 message BertResInstance { repeated Embedding_values instances = 1; };
 
-message Response { repeated BertResInstance instances = 1; };
+message Response {
+  repeated BertResInstance instances = 1;
+  optional int64 op_time = 2;
+  optional int64 infer_time = 3;
+};
 
 service BertService {
   rpc inference(Request) returns (Response);