> ## Documentation Index
> Fetch the complete documentation index at: https://mintlify.com/helicone/helicone/llms.txt
> Use this file to discover all available pages before exploring further.

# Create Evaluation

> Create a new evaluation score for a specific request

This endpoint allows you to add an evaluation score to a specific request. Use this to track quality metrics, performance scores, or custom evaluation criteria for your LLM requests.

## Use Cases

* Add human feedback scores to requests
* Record automated evaluation results
* Track custom quality metrics
* Build evaluation datasets for model improvements

## Path Parameters

<ParamField path="requestId" type="string" required>
  The unique identifier of the request to evaluate. This is the Helicone request ID returned when logging requests.
</ParamField>

## Request Body

<ParamField body="name" type="string" required>
  Name of the evaluation metric (e.g., "accuracy", "relevance", "quality")
</ParamField>

<ParamField body="score" type="number" required>
  Numerical score for the evaluation. Can be any number, but typically normalized to a range like 0-1 or 0-100.
</ParamField>

## Response

Returns a Result object indicating success or failure.

<ResponseField name="data" type="null">
  Null on success
</ResponseField>

<ResponseField name="error" type="string | null">
  Error message if the request failed, null on success
</ResponseField>

## Example Request

```bash theme={null}
curl --request POST \
  --url https://api.helicone.ai/v1/evals/550e8400-e29b-41d4-a716-446655440000 \
  --header 'Authorization: Bearer <YOUR_API_KEY>' \
  --header 'Content-Type: application/json' \
  --data '{
    "name": "accuracy",
    "score": 0.95
  }'
```

## Example Response

```json theme={null}
{
  "data": null,
  "error": null
}
```

## Error Response

```json theme={null}
{
  "data": null,
  "error": "Request not found"
}
```

## Best Practices

* Use consistent evaluation metric names across your requests
* Normalize scores to a standard range (e.g., 0-1)
* Add evaluations soon after request completion for accurate tracking
* Consider using multiple evaluation metrics to capture different quality dimensions