Evaluator Stats

Get statistics and performance metrics for an evaluator

Endpoint

method · string · default: "GET"

GET

endpoint · string

/v1/evaluator/{evaluatorId}/stats

Authentication

This endpoint requires API key authentication. Include your API key in the request headers:

Authorization: Bearer YOUR_API_KEY

Path Parameters

evaluatorId · string · required

The unique identifier of the evaluator

Response

data · object

Evaluator statistics

averageScore · number

Average score across all evaluations

totalUses · number

Total number of times this evaluator has been used

recentTrend · string

Recent trend direction: “up”, “down”, or “stable”

scoreDistribution · array

Distribution of scores across ranges

range · string

Score range (e.g., “0-20”, “20-40”)

count · number

Number of evaluations in this range

timeSeriesData · array

Historical data over time

date · string

Date in ISO format

value · number

Average score for that date

error · string | null

Error message if the request failed, null otherwise

Example Request

curl -X GET https://api.helicone.ai/v1/evaluator/eval_abc123/stats \
  -H "Authorization: Bearer YOUR_API_KEY"

Example Response

{
  "data": {
    "averageScore": 7.5,
    "totalUses": 1250,
    "recentTrend": "up",
    "scoreDistribution": [
      {
        "range": "0-2",
        "count": 50
      },
      {
        "range": "2-4",
        "count": 100
      },
      {
        "range": "4-6",
        "count": 200
      },
      {
        "range": "6-8",
        "count": 450
      },
      {
        "range": "8-10",
        "count": 450
      }
    ],
    "timeSeriesData": [
      {
        "date": "2024-01-01",
        "value": 7.2
      },
      {
        "date": "2024-01-02",
        "value": 7.3
      },
      {
        "date": "2024-01-03",
        "value": 7.4
      },
      {
        "date": "2024-01-04",
        "value": 7.6
      },
      {
        "date": "2024-01-05",
        "value": 7.5
      }
    ]
  },
  "error": null
}

Use Cases

• Monitor evaluator performance over time
• Identify trends in evaluation scores
• Understand score distribution patterns
• Track usage frequency of different evaluators
• Optimize evaluation criteria based on historical data