Update main.py

main.py

@@ -473,7 +473,7 @@ async def chat(request: ChatRequest):
 
         # Use token from request or fallback to env
         token = request.hf_token or hf_token
-
+        const hf_token = os.getenv("HUGGINGFACE_TOKEN")
         # Generate response
         if not token:
             response = f"""[LLM Response Placeholder]
@@ -494,7 +494,7 @@ Example:
         else:
             try:
                 client = InferenceClient(
-                    token=token,
+                    token=hf_token,
                     model="openai/gpt-oss-20b"
                 )