llm.ReasoningEffort

Note:

The content in this help topic pertains to SuiteScript 2.1.

Enum Description

The reasoning effort to use for LLM requests.

Reasoning effort controls how much internal reasoning a model uses before providing a response. Higher reasoning effort can improve the quality of the response on complex tasks but may increase token usage (and NetSuite AI Unit usage) and response time. You should choose the lowest value that meets your response quality requirements.

Use this enum to set the value of the options.modelParameters.reasoningEffort parameter in llm.generateText(options) and llm.generateTextStreamed(options).

Reasoning effort is available only for GPT OSS models.

Note:

JavaScript does not include an enumeration type. The SuiteScript 2.x documentation uses the term enumeration (or enum) to describe a plain JavaScript object with a flat, map-like structure. In this object, each key points to a read-only string value.

Module

N/llm Module

Supported Script Types

Server scripts

For more information, see SuiteScript 2.1 Script Types.

Since

2026.2

Values

Enum	Value	Notes
`ReasoningEffort.LOW`	`LOW`	Low reasoning effort uses minimal reasoning to prioritize faster responses and lower token use. It's suitable for straightforward tasks and time-sensitive workflows.
`ReasoningEffort.MEDIUM`	`MEDIUM`	Medium reasoning effort provides a balanced level of reasoning, quality, timeliness, and token use. It's suitable as a general-purpose setting for tasks that require some analysis or multiple steps. This value is the default value when sending requests to GPT OSS models using llm.generateText(options) and llm.generateTextStreamed(options).
`ReasoningEffort.HIGH`	`HIGH`	High reasoning effort uses more thorough reasoning for complex problems where quality is more important than response time. It's suitable for difficult analysis, multi-step planning, and tasks that require careful evaluation of information.

Enum

Value

Notes

ReasoningEffort.LOW

LOW

Low reasoning effort uses minimal reasoning to prioritize faster responses and lower token use. It's suitable for straightforward tasks and time-sensitive workflows.

ReasoningEffort.MEDIUM

MEDIUM

Medium reasoning effort provides a balanced level of reasoning, quality, timeliness, and token use. It's suitable as a general-purpose setting for tasks that require some analysis or multiple steps.

This value is the default value when sending requests to GPT OSS models using llm.generateText(options) and llm.generateTextStreamed(options).

ReasoningEffort.HIGH

HIGH

High reasoning effort uses more thorough reasoning for complex problems where quality is more important than response time. It's suitable for difficult analysis, multi-step planning, and tasks that require careful evaluation of information.

Syntax

Important:

The following code sample shows the syntax for this member. It isn't a functional example. For a complete script example, see N/llm Module Script Samples.

            // Add additional code
...

const response = llm.generateText({
   prompt: "Hello world!",
   reasoningEffort: llm.ReasoningEffort.LOW
});

...
// Add additional code

General Notices