ReasoningEffortLevels

Specifies the level of reasoning effort the model should apply when generating a response.

Remarks

The ReasoningEffortLevels enumeration allows you to control how much computational effort and depth of reasoning the model should use when producing a response. This setting can influence the quality, accuracy, and latency of completions, especially for complex or ambiguous tasks.

Usage Guidance

• Use lower levels for simple, fast responses where minimal reasoning is required.
• Use higher levels for tasks that demand careful analysis, multi-step reasoning, or higher accuracy.
• Not all models or service tiers may support all levels; refer to the OpenAI API documentation for availability.

Impact

Increasing the reasoning effort may improve the quality and reliability of answers, but can also increase response time and cost.

Can be one of the following values:

Used By

Chat Parameters