CustomEvalRequestSchema

model_1_namestringrequired

Name of the primary model whose outputs will be evaluated.

Example: Llama-3.3-70B-Instruct

model_2_namestringrequired

Name of the baseline comparison model.

Example: Mistral-Small-24B-Instruct-2501

custom_eval_namestring

Name of the custom evaluation definition to use for scoring, created via Create custom evaluation.

Default: null

Example: example-eval

CustomEvalRequestSchema

{
  "model_1_name": "Llama-3.3-70B-Instruct",
  "model_2_name": "Mistral-Small-24B-Instruct-2501",
  "custom_eval_name": "example-eval"
}