feat: reduce thinking budget to lower response time

app/agent.py

@@ -73,7 +73,8 @@ def create_response(ctx: RunContext, output: list[Statement]) -> models.Statemen
 
 model = GoogleModel("gemini-2.5-flash-preview-05-20")
 settings = GoogleModelSettings(
-    google_thinking_config={"thinking_budget": 2048, "include_thoughts": True},
+    # temperature=0.1,
+    google_thinking_config={"thinking_budget": 1024, "include_thoughts": True},
 )
 agent = Agent(
     model=model,