diff --git a/utils/api_models.py b/utils/api_models.py
index d3edcc1..f1a8044 100644
--- a/utils/api_models.py
+++ b/utils/api_models.py
@@ -5,7 +5,7 @@ API data models and response schemas.
 
 from typing import Dict, List, Optional, Any, AsyncGenerator
 from pydantic import BaseModel, Field, field_validator, ConfigDict
-
+from utils.settings import DEFAULT_THINKING_ENABLE
 
 class Message(BaseModel):
     role: str
@@ -53,7 +53,7 @@ class ChatRequest(BaseModel):
     robot_type: Optional[str] = "general_agent"
     user_identifier: Optional[str] = ""
     session_id: Optional[str] = None
-    enable_thinking: Optional[bool] = True
+    enable_thinking: Optional[bool] = DEFAULT_THINKING_ENABLE
 
 
 class ChatRequestV2(BaseModel):
@@ -64,7 +64,7 @@ class ChatRequestV2(BaseModel):
     language: Optional[str] = "zh"
     user_identifier: Optional[str] = ""
     session_id: Optional[str] = None
-    enable_thinking: Optional[bool] = True
+    enable_thinking: Optional[bool] = DEFAULT_THINKING_ENABLE   
 
 
 class FileProcessRequest(BaseModel):
diff --git a/utils/settings.py b/utils/settings.py
index 0123613..f22492c 100644
--- a/utils/settings.py
+++ b/utils/settings.py
@@ -37,3 +37,6 @@ SENTENCE_TRANSFORMER_MODEL = os.getenv("SENTENCE_TRANSFORMER_MODEL", "TaylorAI/g
 # Tool Output Length Control Settings
 TOOL_OUTPUT_MAX_LENGTH = int(SUMMARIZATION_MAX_TOKENS/4)
 TOOL_OUTPUT_TRUNCATION_STRATEGY = os.getenv("TOOL_OUTPUT_TRUNCATION_STRATEGY", "smart")
+
+# THINKING ENABLE
+DEFAULT_THINKING_ENABLE = os.getenv("DEFAULT_THINKING_ENABLE", "true") == "true"