4.1-mini

Author: bcherry

tests/test_agent.py

@@ -6,8 +6,7 @@
 
 def _judge_llm() -> llm.LLM:
     # We can use a different LLM to evaluate the agent's responses than the one used in the agent itself
-    # This allows you to use reasoning capabilities or larger models than would be practical for realtime chat
-    return inference.LLM(model="openai/gpt-5.2")
+    return inference.LLM(model="openai/gpt-4.1-mini")
 
 
 @pytest.mark.asyncio