EvalCandidateResolved

Fields

What the run ACTUALLY replays (stored in ai_eval_runs.candidate).

6 fields

model

string or null

offline_mode · required

boolean

prompt_version

string or null

provider

string or null

requested · required

EvalCandidateIn

tier · required

string

Raw schema (JSON)

{
  "description": "What the run ACTUALLY replays (stored in ``ai_eval_runs.candidate``).",
  "properties": {
    "model": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "title": "Model"
    },
    "offline_mode": {
      "title": "Offline Mode",
      "type": "boolean"
    },
    "prompt_version": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "title": "Prompt Version"
    },
    "provider": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "title": "Provider"
    },
    "requested": {
      "$ref": "#/components/schemas/EvalCandidateIn"
    },
    "tier": {
      "title": "Tier",
      "type": "string"
    }
  },
  "required": [
    "tier",
    "offline_mode",
    "requested"
  ],
  "title": "EvalCandidateResolved",
  "type": "object"
}

Used by endpoints

4 endpoints send or return this schema, on its own or inside another schema.

• GET /api/v1/ai-eval/dashboard — Get Quality Dashboard
• GET /api/v1/ai-eval/runs — List Eval Runs
• GET /api/v1/ai-eval/runs/{eval_run_id} — Get Eval Run
• POST /api/v1/ai-eval/runs/evaluate/{dataset_id} — Run Evaluation

Used by schemas

• AIEvalRunResponse

More details

• Specification: OpenAPI JSON