Skip to main content
Environment Node Node inspector for an LLM Evaluator node

​
Parameters

​
Model Name
string
default:"openai/gpt-5-mini"
required
LLM used to evaluate the run.
​
API Version
string
Optional provider API version to use when calling the model.
​
Reasoning Effort
string
The level of reasoning effort to use for the LLM, if supported by the model. Accepted values are none, minimal, low, medium, high, xhigh, or default.
​
Extra Model Arguments
json
Additional arguments to pass directly to the underlying model’s API.
​
System Prompt
string
required
Instructions for the evaluator. You can use expressions like ${input} to reference the agent input and ${@NodeA.output} to reference upstream node outputs.
​
Evaluator Input
string
required
Input passed to the evaluator. Supports the agent input editor. You can use expressions like ${input} to reference the agent input and ${@NodeA.output} to reference upstream node outputs.

​
Credentials

​
LLM Credential
LLM Credential
required
Select an LLM Credential used by the evaluator model.

​
Inputs

​
Input
any
The input passed to the evaluator. This can be used for evaluation by using an expression like ${input} in the Evaluator Input parameter.

​
Outputs

​
Output
json
Evaluation result with score and feedback.
FieldDescription
scoreNumeric score for the evaluation.
feedbackNatural language feedback for the evaluation.