Free tier — no credit card

Agent governance for
every AI tool. Free.

Bring your own API keys and get the full Clevername platform at no cost — agent governance, runtime enforcement, MCP Gateway, and Credential Vault. Your AI provider charges you directly; Clevername never marks up LLM costs.

Free tier

Everything included at $0

Bring your own API keys and use any model, unlimited. We handle the guardrails — you pay your provider directly.

Unlimited BYOK RoutingBYOK only
Connect your own API keys from Anthropic, OpenAI, or Google. Unlimited requests, any model. You pay your provider directly — we never mark up LLM costs.
CleverGuard — All 5 Tiers
Data label gate, pattern detection, ML injection classifier, NER entity recognition, and LLM semantic analysis. Every tier included on every plan, always on.
Agent Review
27-question security intake, guardrail profile compilation, and governance gates for agent deployment. Included on all plans.
AI Company — Full Platform
Create agents, define roles and tools, compile guardrail profiles. Up to 5 governed agents (with gateway tokens) on Free.
MCP Gateway + Integrations
Unlimited MCP server installs. Scoped gateway tokens with tool allowlists and spend budgets — one per governed agent.
Credential Vault
API keys stored in GCP Secret Manager — never in our database. Rotate or revoke at any time without touching your code.
Audit Logging
Every API call logged with tamper-evident hash chain. 7-day retention on Free.
Get started free — no credit card

No credit card required. Full platform, forever free with BYOK.

CleverGuard Scanning

Five tiers of protection. Included on all plans.

Every scan passes through all five security tiers automatically. Higher tiers only fire when earlier tiers need a second opinion — keeping latency low for clean traffic.

Tier 0
All plans

Data Label Gate

Deterministic rank check · <1µs

Enterprise sensitivity labels (Purview, Google DLP, AWS Macie). Blocks before any content analysis if data classification exceeds agent clearance.

Tier 1
All plans

Pattern Shield

Regex + 11-step evasion decoding · <1ms

Catches injection patterns, ROT13/base64 obfuscation, Unicode homoglyphs, BiDi overrides, Zalgo text, zero-width chars, and credential leaks.

Tier 1.5
All plans

ML Classifier

Prompt Guard 2 (86M params) · ~15ms

Semantic injection detection via fine-tuned transformer. Resistant to character-level mutation attacks that bypass regex. Input direction only.

Tier 2
All plans

Entity Recognition

Presidio NER inference · ~50ms

Finds names, SSNs, emails, credit cards in natural language context. Entity types vary by compliance framework (HIPAA, PCI, GDPR).

Tier 3
All plans

AI Deep Scan

LLM semantic analysis (BYOK) · ~2s

Full intent analysis with agent context. Catches social engineering, subtle manipulation, and novel attack patterns. Only triggered when needed. Uses your API key.

Tiers 0 — 2Every requestData labels, regex, ML classifier, NER — all plans, always on
Tier 3 — AI Deep ScanOn demandFires when earlier tiers flag ambiguous content. Uses your BYOK key.
Typical L3 trigger rate<5%Most traffic never touches Layer 3
Paid plans

Upgrade when you need more

Free gives you the full platform. Pro unlocks deeper scanning and higher limits for $29/mo. Enterprise adds multi-seat governance, SSO, and dedicated support.

Pro

$29/ month

for individual power users

Everything in Free, plus:

  • +15 governed agents + gateway tokens (expandable)
  • +15 guardrail profiles
  • +30-day audit log retention
  • +Data never used for training
  • +Email support
Go Pro

Enterprise

Contact us

Everything in Pro, built for your security org. Multi-seat, volume pricing, SIEM forwarding, SLA, and a human who answers the phone.

Get in touch
+Unlimited seats, governed agents, and gateway tokens
+Cloud or self-hosted deployment (same Docker image)
+BYOL — log export to your SIEM (Splunk, Datadog, Sentinel)
+Org custom scanner rules — block patterns, URL lists, entity definitions
+Custom audit log retention
+99.95% uptime SLA with dedicated oncall
+Dedicated support channel
+Annual license — CapEx-friendly, no usage fees when self-hosted
Cost Transparency

What it actually costs us

We publish our costs because we want you to trust the pricing. No hidden multipliers. BYOK users pay nothing to us — ever.

WhatOur costYour cost
Layer 1 — Pattern Shield
~$0.000002 / scanFree
Layer 2 — Entity Recognition
~$0.000005 / scanFree
Layer 3 — Semantic (BYOK)
$0 to usYour provider bill
LLM chat (BYOK)
$0 to usYour provider bill
~$0.12Max cost per free BYOK user / monthL1+L2 scanning included. L3 costs us nothing on BYOK.
<50msStandard scan — L1 + L2 alwaysPattern matching + Presidio NER. Runs on every request. Always included.
<5%L3 trigger rateAI Deep Scan only fires on ambiguous content. BYOK uses your own key.
Plan comparison

What each plan gets you

FeatureFreeProEnterprise
CleverGuard ScanningAll 5 tiersAll 5 tiersAll 5 tiers + org custom rules
Governed agents + gateway tokens515 (expandable)Unlimited
Guardrail profiles515Unlimited
Audit retention7 days30 daysCustom
AI Company + Agent ReviewFullFullFull
LLM Gateway (BYOK)UnlimitedUnlimitedUnlimited
Data used for trainingYesNeverNever
SSO + SCIMIncluded
SOC dashboardIncluded
Org policy controlsIncluded
SIEM forwardingIncluded
Org membersUp to 3Up to 3Unlimited
All plans include

CleverGuard — all 5 tiers, free for everyone

Data label enforcement, regex + evasion decoding, ML injection classifier, NER entity detection, and LLM semantic analysis. Every tier included on every plan.

See what it catches
FAQ

Common questions