Agent governance for AI tools.
Connect your agents and MCP servers to Clevername — we route every call and scan it before it reaches the model. You pay your AI provider directly; Clevername never marks up LLM costs.
Plans that scale with you
Every plan gets the full Clevername scanning stack. Paid tiers buy more seats, more agents, longer retention, and SSO — not more security.
Free
$0Connect & scan · individuals
- +Connect any provider, unlimited requests
- +Full scanning stack, all tiers
- +5 governed agents + gateway tokens
- +MCP Gateway, unlimited installs
- +Credential Vault
- +7-day audit retention
Pro
for individual power users
- +25 governed agents — expand with $10/mo +5 agent packs
- +30-day audit log retention
- +5 webhook endpoints
- +Email support
Team
2-seat minimum · per-seat
- +Up to 25 seats, 25 governed agents — expand with $10/mo +5 agent packs
- +365-day audit log retention
- +SOC dashboard — CleverGuard monitoring
- +Agent Review board — multi-member council
- +Org-wide policy controls + custom scanner rules
- +Priority support
Enterprise
Contact usBuilt for security orgs
- +Unlimited seats, agents, and gateway tokens
- +SSO + SCIM directory sync
- +Cloud or self-hosted, same image
- +SIEM log export (Splunk, Datadog, Sentinel)
- +Dedicated support channel
Subscriptions renew monthly. Cancel anytime from your billing settings — no cancellation fee. Downgrades take effect at the end of the current billing period. Unused time is not refunded; upgrades are prorated to the day.
Tier-3 deep-scan is Clevername-operated self-hosted governance compute; there is no customer Tier-3 key path or managed-provider fallback. It is wired for governed production traffic, still labeled beta, and surfaces reduced coverage if unavailable.
Every plan includes the full guard suite — CleverGuard, ScopeGuard, ClaimGuard, DriftGuard — routed through MCP Gateway. No tier is paywalled; paid plans buy seats and retention, not protection. (CleverGuard Tier-3 deep-scan is self-hosted beta governance compute — see below.)
See what each one catches →What each plan gets you
| Feature | Free | Pro | Team | Enterprise |
|---|---|---|---|---|
| Traffic Security — scanned on every call | ||||
| CleverGuard scanning (T0–T3*) | All 4 tiers* | All 4 tiers* | All 4 tiers* + org custom rules | All 4 tiers* + org custom rules |
| Credential interception + PII detection | ✓ included | ✓ included | ✓ included | ✓ + KMS BYOK |
| Audit log retention | 7 days | 30 days | 1 year | Custom |
| LLM Gateway (BYOK) | Unlimited | Unlimited | Unlimited | Unlimited |
| Personal access keys (IDE / scripts) | Unlimited | Unlimited | Unlimited | Unlimited |
| Agent Governance — governed agents only | ||||
| Governed agents + CNK tokens | 5 | 25 (+ packs) | 25 (+ packs) | Unlimited |
| Agent Review + guardrail profiles | ✓ self-approve | ✓ self-approve | ✓ + multi-member council | ✓ + bulk + SSO |
| ScopeGuard — mandate enforcement | ✓ included | ✓ included | ✓ + custom rules | ✓ + custom rules |
| ClaimGuard — hallucination checks | ✓ included | ✓ included | ✓ + custom config | ✓ + custom config |
| DriftGuard — behavioral drift | ✓ included | ✓ included | ✓ included | ✓ + SIEM forward |
| SSO + SCIM sync | — | — | — | ✓ included |
| Deployment | ||||
| Cloud (managed) | ✓ | ✓ | ✓ | ✓ |
| Self-hosted | — | — | — | ✓ same Docker image |
| Data used for training | Never | Never | Never | Never |
| Org members | Just you | Just you | Up to 25 | Unlimited |
* Tier-3 deep-scan is self-hosted beta governance compute, wired for governed production traffic. Tiers 0–2 remain the always-on baseline.