Compare Ai With Baseline — GET /api/v1/runs/{run_id}/ai-compare
Compare Ai With Baseline
This run’s AI analysis against a baseline’s (default: the D42 previous run). Live sides by default; a pinned version compares what a pipeline produced. Never triggers an LLM.
Endpoint: GET /api/v1/runs/{run_id}/ai-compare
Authentication: Send a signed-in user’s access token in Authorization: Bearer <access token>, or an API key in X-API-Key: <API key>.
- You must have access to this test run.
Parameters
Section titled “Parameters”| Name | In | Required | Type | Default | Description |
|---|---|---|---|---|---|
run_id | path | yes | string (uuid) | ||
baseline_run_id | query | no | string (uuid) or null | ||
suite_name | query | no | string or null | ||
version_id | query | no | string (uuid) or null | A digest version id of THIS run | |
baseline_version_id | query | no | string (uuid) or null | A digest version id of the BASELINE run | |
X-API-Key | header | no | string or null | API key, as an alternative to the access token (see Authentication). |
Request body
Section titled “Request body”No request body.
Responses
Section titled “Responses”| Status | Meaning | Body |
|---|---|---|
200 | Successful Response | AICompareResponse (application/json) |
422 | Validation Error | HTTPValidationError (application/json) |
Errors this endpoint returns, read from its source code:
422, with the message “baseline_run_id must be a different run”
It can also return 401 if your credential is missing or expired, 403 if your credential does not allow this request, and 404 if the item does not exist or you cannot see it. Your role, your project membership and the key’s scope decide what a credential allows. See Errors and retries.
More details
Section titled “More details”- Group: AI History
- Operation ID:
compare_ai_with_baseline_api_v1_runs__run_id__ai_compare_get - Source code:
backend/app/routers/ai_history.py, line 253 - Specification: OpenAPI JSON