diff --git a/api/simple_chat.py b/api/simple_chat.py
index 9076f71ab..749d2ebe6 100644
--- a/api/simple_chat.py
+++ b/api/simple_chat.py
@@ -11,7 +11,7 @@
 from api.chat import ChatStreamer, prompt_builder, is_token_limit_error
 from api.config import get_model_config, configs
 from api.data_pipeline import count_tokens, get_file_content
-from api.rag import RAG
+from api.rag import RAG, MAX_INPUT_TOKENS
 from api.prompts import (
     DEEP_RESEARCH_FIRST_ITERATION_PROMPT,
     DEEP_RESEARCH_FINAL_ITERATION_PROMPT,
@@ -78,8 +78,8 @@ async def chat_completions_stream(request: ChatCompletionRequest):
             if hasattr(last_message, 'content') and last_message.content:
                 tokens = count_tokens(last_message.content, request.provider == "ollama")
                 logger.info(f"Request size: {tokens} tokens")
-                if tokens > 8000:
-                    logger.warning(f"Request exceeds recommended token limit ({tokens} > 7500)")
+                if tokens > MAX_INPUT_TOKENS:
+                    logger.warning(f"Request exceeds recommended token limit ({tokens} > {MAX_INPUT_TOKENS})")
                     input_too_large = True
 
         # Create a new RAG instance for this request
diff --git a/api/websocket_wiki.py b/api/websocket_wiki.py
index 2d7e45b71..2b7d17b85 100644
--- a/api/websocket_wiki.py
+++ b/api/websocket_wiki.py
@@ -13,7 +13,7 @@
     configs,
 )
 from api.data_pipeline import count_tokens, get_file_content
-from api.rag import RAG
+from api.rag import RAG, MAX_INPUT_TOKENS
 from api.prompts import (
     DEEP_RESEARCH_FIRST_ITERATION_PROMPT,
     DEEP_RESEARCH_FINAL_ITERATION_PROMPT,
@@ -75,8 +75,8 @@ async def handle_websocket_chat(websocket: WebSocket):
             if hasattr(last_message, 'content') and last_message.content:
                 tokens = count_tokens(last_message.content, request.provider == "ollama")
                 logger.info(f"Request size: {tokens} tokens")
-                if tokens > 8000:
-                    logger.warning(f"Request exceeds recommended token limit ({tokens} > 7500)")
+                if tokens > MAX_INPUT_TOKENS:
+                    logger.warning(f"Request exceeds recommended token limit ({tokens} > {MAX_INPUT_TOKENS})")
                     input_too_large = True
 
         # Create a new RAG instance for this request