Per organisation, billed monthly. Test usage is included in every plan; live actions and AI-backed checks are metered transparently and shown per run.
AI usage (semantic judge, guardrail judges, Test Architect, red team) is metered at cost — bring your own key or use the managed gateway. Deterministic checks never need AI.
Capabilities, not honesty. Every plan gets the same deterministic verdicts.
We enable Guard on a Free workspace for pilots — ask from Settings → Plan.
Talk to usOne suite run — any number of scenarios × repeats. Free includes 100 per month, Test 2,000, Guard unlimited.
No. Deterministic checks decide every verdict. An AI key (yours or managed) adds the advisory semantic judge, guardrail judges, the Test Architect and AI-generated red-team suites — all metered and never €0.
Yes, as an entitlement: request it from Settings → Plan and a COLVO admin can enable it for a pilot. The plan stays the baseline.
Per call, per run and per month, as known (gateway-priced) or estimated (from the model catalog). Caps at org and platform level stop calls at 100 % and email you at 80 %.