While using groq models we get the thinking section added with actual llm response. There we can use reasoning_format="parsed" . Check below Without reasoning_format llm…