Pricing that scales with your governance
Start free, upgrade when AI moves to production. Every plan includes the full audit trail.
Starter
For individuals exploring AI validation
$0
- Up to 5 models in the registry
- 10 validation runs / month
- Community test suites
- Use Case Advisor recommendations
- 1 workspace seat
Validation compute, metered fairly
Pay per gate execution beyond your plan’s included runs. No hidden seat minimums.
| Suite | Coverage | Billed per | Price |
|---|---|---|---|
| Security & Adversarial suite | Prompt injection, jailbreak and exploit probes | run | $0.40 / run |
| Reliability & Grounding suite | Hallucination and groundedness checks | run | $0.25 / run |
| Compliance & Regulatory suite | Framework-mapped control verification | run | $0.35 / run |
| Performance & Latency suite | Latency SLO and throughput regression gates | run | $0.15 / run |
| Custom probe pack | Bring-your-own adversarial probe corpus | 1k probes | $0.10 / 1k probes |
Frequently asked
- What counts as a validation run?
- One run executes a full gate — every check in the selected suite against one model version. Partial suite executions are billed per completed check.
- Can I self-host the validation engine?
- Yes. Enterprise workspaces can run the engine inside their own VPC or on-prem cluster with the same evidence and audit guarantees.
- Do unused runs roll over?
- Team-plan run quotas reset monthly. Enterprise plans can negotiate pooled or committed-use pricing.
- Is there a discount for startups or academia?
- Early-stage startups and research institutions get 50% off the Team plan for the first year — mention it during onboarding.