Download Evaluation Results

Download evaluation results as a JSON file.

This endpoint returns the complete evaluation results as a downloadable JSON file. The file includes all metrics, scores, and detailed evaluation data.

Path Parameters:

• evaluation_id: Unique evaluation identifier

Returns:

JSON file download containing:

• Single model evaluation: All metrics, scores, and detailed results
• Comparison mode: Both models’ metrics, scores, and comparison data

Response Headers:

• Content-Type: application/json
• Content-Disposition: attachment; filename=evaluation_results_{evaluation_id}.json or evaluation_comparison_{evaluation_id}.json

Raises:

• 404: If evaluation not found or doesn’t belong to the authenticated user
• 400: If evaluation is not yet completed

Example Request:

GET /api/v3/evaluations/eval-xyz789/download
Headers: {"Authorization": "Bearer <api_key>"}

Example Downloaded File (Single Model):

{
  "evaluation_id": "eval-xyz789",
  "status": "completed",
  "model_id": "claude-3-sonnet-20240229",
  "benchmark_id": "task-abc123",
  "metrics": {
    "accuracy": 0.92,
    "relevance": 0.88,
    "average_score": 0.90
  },
  "detailed_results": [
    {
      "question": "What is the main purpose?",
      "expected_answer": "To provide...",
      "model_answer": "The main purpose is to provide...",
      "score": 0.95
    }
  ],
  "created_at": "2024-01-15T10:30:00Z",
  "completed_at": "2024-01-15T10:45:00Z"
}

Notes:

• Returns formatted JSON with proper indentation
• Filename automatically includes evaluation ID
• Only available for completed evaluations