Test assessments for building an integration
Create an assessment with test: true and it runs the real candidate flow, sends the real webhooks and returns a realistic result, without using a credit.
Building an integration means creating a lot of assessments that nobody will ever be hired from. Those should not cost anything, and they should behave exactly like the real thing, because the parts of an integration that break are webhook delivery, result parsing and status changes.
Test assessments do that. Everything runs for real except the scoring, which returns a synthetic result instead. You can ask for a particular CEFR level, so you can check how your system handles a weak candidate and a strong one.
Synthetic results are labelled everywhere they appear, in the API response, the webhook and the feedback text, so they can never be mistaken for a real score. They are left out of your usage and reporting, and they work even when an account has run out of credit.