reasoningEffort

Constrains effort on reasoning for reasoning models. Reduces reasoning in responses, which can reduce latency and cost at the expense of quality.

Accepts values: "low", "medium", or "high".