diff --git a/README.md b/README.md
index 8ec76ef..d9a1c96 100644
--- a/README.md
+++ b/README.md
@@ -137,6 +137,10 @@ Whether to use Elasticsearch, default is `True`.
 
 The connection arguments to connect elastic service.
 
+***token_model***
+
+The model used to count tokens, defaults to 'gpt-3.5-turbo'.
+
 <br />
 
 
diff --git a/osschat_insert.py b/osschat_insert.py
index 8007b7e..5e69af5 100644
--- a/osschat_insert.py
+++ b/osschat_insert.py
@@ -40,6 +40,8 @@ class OSSChatInsertConfig(BaseModel, extra=Extra.allow):
     # config for elasticsearch
     es_enable: Optional[bool] = True
     es_connection_kwargs: Optional[dict] = {'hosts': ['https://127.0.0.1:9200'], 'basic_auth': ('elastic', 'my_password')}
+    # token count
+    token_model: Optional[str] = 'gpt-3.5-turbo'
 
 
 _hf_models = ops.sentence_embedding.transformers().get_op().supported_model_names()
@@ -96,6 +98,7 @@ def osschat_insert_pipe(config):
     p = (
         pipe.input('doc', 'project_name')
             .map('doc', 'text', data_loader)
+            .map('doc', 'token_count', ops.token_counter(config.token_model))
             .flat_map('text', 'chunk', text_split_op)
             .map('chunk', 'embedding', sentence_embedding_op, config=sentence_embedding_config)
     )
@@ -109,8 +112,8 @@ def osschat_insert_pipe(config):
         p = (
             p.map('chunk', 'es_doc', lambda x: {'doc': x})
              .map(('project_name', 'es_doc'), 'es_res', es_index_op)
-             .map(('milvus_res', 'es_res'), 'res', lambda x, y: {'milvus_res': x, 'es_res': y})
+             .map(('milvus_res', 'es_res', 'token_count'), 'res', lambda x, y, c: {'milvus_res': x, 'es_res': y, 'token_count': c})
         )
     else:
-        p = p.map('milvus_res', 'res', lambda x: {'milvus_res': x, 'es_res': None})
+        p = p.map(('milvus_res', 'token_count'), 'res', lambda x, c: {'milvus_res': x, 'es_res': None, 'token_count': c})
     return p.output('res')