diff --git a/fastdeploy/engine/common_engine.py b/fastdeploy/engine/common_engine.py
index 41c12a2d6..e6b8442a1 100644
--- a/fastdeploy/engine/common_engine.py
+++ b/fastdeploy/engine/common_engine.py
@@ -47,7 +47,7 @@ from fastdeploy.splitwise.splitwise_connector import SplitwiseConnector
 from fastdeploy.utils import EngineError, envs, llm_logger
 
 
-class EngineSevice:
+class EngineService:
     """
     Base class containing common engine functionality
     """
diff --git a/fastdeploy/engine/engine.py b/fastdeploy/engine/engine.py
index a71a2df61..a710d8cc9 100644
--- a/fastdeploy/engine/engine.py
+++ b/fastdeploy/engine/engine.py
@@ -34,7 +34,7 @@ import paddle
 from tqdm import tqdm
 
 from fastdeploy.engine.args_utils import EngineArgs
-from fastdeploy.engine.common_engine import EngineSevice
+from fastdeploy.engine.common_engine import EngineService
 from fastdeploy.engine.expert_service import start_data_parallel_service
 from fastdeploy.engine.request import Request
 from fastdeploy.input.preprocess import InputPreprocessor
@@ -93,7 +93,7 @@ class LLMEngine:
             cfg.model_config.enable_mm,
             cfg.tool_parser,
         )
-        self.engine = EngineSevice(cfg)
+        self.engine = EngineService(cfg)
 
         if self.cfg.cache_config.num_gpu_blocks_override is None:
             self.do_profile = 1
diff --git a/fastdeploy/engine/expert_service.py b/fastdeploy/engine/expert_service.py
index 87a09b96f..662cd58f7 100644
--- a/fastdeploy/engine/expert_service.py
+++ b/fastdeploy/engine/expert_service.py
@@ -25,7 +25,7 @@ import weakref
 
 import numpy as np
 
-from fastdeploy.engine.common_engine import EngineSevice
+from fastdeploy.engine.common_engine import EngineService
 from fastdeploy.inter_communicator import IPCSignal
 from fastdeploy.utils import console_logger, envs, llm_logger
 
@@ -64,7 +64,7 @@ class ExpertService:
             else:
                 self.cfg.cache_config.pd_comm_port = [self.cfg.cache_config.pd_comm_port[local_data_parallel_id]]
         self.cfg.parallel_config.local_data_parallel_id = local_data_parallel_id
-        self.engine = EngineSevice(self.cfg, start_queue)
+        self.engine = EngineService(self.cfg, start_queue)
         if self.cfg.scheduler_config.name == "splitwise":
             self.engine.scheduler.reset_nodeid(f"{self.engine.scheduler.infer.nodeid}_{local_data_parallel_id!s}")