Checking API health.

Log in

Judge

Eval runs

Launch a spec against a dataset. The API scores the rows and this page polls until the status is no longer active.

No specs yet

Compile a draft before launching a run.

No datasets yet

Import rows, then come back to score them.

Leave blank to score every row. A name such as test scores only that split.

No runs yet

Launch a spec against a dataset to score its rows.