POST /v1/eval
Run an evaluation by providing a saved evaluation spec and a dataset of input/output pairs.Headers
Authorization: Api-Key: API key for authentication
Request body
eval_spec_id(string): Identifier of the evaluation spec to usedataset(array): List of objects with:input(string): Prompt or expected inputoutput(string): Model output to grade
model(string, optional): Override the default model used for evaluation
Example request
Example response
Status codes
200 OK– Evaluation completed400 Bad Request– Missing or empty dataset403 Forbidden– API key not authorized for the evaluation spec404 Not Found– Evaluation spec or version not found
GET /v1/eval//documents
Retrieve the processed documents that belong to a dataset, including presigned URLs for downloading the original files.Path parameters
dataset_id(string): Identifier of the dataset whose documents you want to download
Headers
Authorization: Api-Key: API key for authentication
Example request
Example response
Status codes
200 OK– Documents were found and returned successfully403 Forbidden– API key is not authorized for the dataset’s organization404 Not Found– Dataset does not exist or it has no documents