Execute Evals for Trace

Queue evaluations to run against a specific trace. Evaluations are executed asynchronously.

Use the eval_types field to control which evaluations run:

• all (default): Run both LLM-as-judge and code evaluations
• llm-as-judge: Run only LLM-based evaluations
• code: Run only code-based evaluations

Optionally pass criteria_ids to run specific evaluation criteria. When provided, those criteria are executed regardless of whether they have already been evaluated. When omitted, only criteria that have not yet been evaluated for this trace will run.

POST /api/v2/projects/{project_id}/sessions/{session_id}/traces/{trace_id}/evaluations

Execute Evals for Trace

cURL

curl --request POST \
  --url https://api.example.com/api/v2/projects/{project_id}/sessions/{session_id}/traces/{trace_id}/evaluations \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '
{
  "eval_types": "all",
  "criteria_ids": null
}
'

202

{}

Authorizations

Authorization · string · header · required

Bearer authentication header of the form Bearer <token>, where <token> is your auth token.

Path Parameters

project_id · string · required

session_id · string · required

trace_id · string · required

Body

application/json

eval_types · enum<string> · default: all

Available options: all, llm-as-judge, code

criteria_ids · string<uuid>[] | null

Response

202 - application/json

EmptyResponse()

The response is of type EmptyResponse · object.