Pricing
Pay for the governance, not the seat.
Start free and put a control plane in front of your AI the same afternoon. Every paid tier governs, routes and proves every call — with a full audit trail your auditors will accept.
Free
Put governance in front of your AI in 5 minutes.
- ✓10K AI operations / month
- ✓1 provider connector
- ✓2 agents
- ✓Basic cost & decision dashboard
- ✓Community support
Starter
Govern, route and prove a single team’s AI.
- ✓100K AI operations / month
- ✓5 provider connectors
- ✓10 agents
- ✓Full decision audit trail
- ✓Budget enforcement & caps
- ✓Prompt Intelligence Engine
- ✓Reused responses (cache)
- ✓AI FinOps command center
- ✓90-day trace retention
- ✓Email support
Growth
Multi-model governance, quality gates and full audit for AI teams.
- ✓5M AI operations / month
- ✓Unlimited connectors & agents
- ✓Everything in Starter
- ✓SSO (OIDC) + RBAC
- ✓Golden-dataset regression tests
- ✓AI quality scoring
- ✓Custom routing & governance policies
- ✓Full signed audit trail
- ✓365-day retention
- ✓Priority SLA
Enterprise
Sovereign AI governance for regulated organisations.
- ✓Unlimited AI operations
- ✓Sovereign / on-premise deployment
- ✓OVH · Scaleway · Outscale compatible
- ✓SAML SSO + advanced RBAC
- ✓SOC 2 — in progress
- ✓GDPR + EU data residency
- ✓Isolated tenant architecture
- ✓Custom audit & compliance reports
- ✓AI risk scoring + EU AI Act readiness
- ✓Custom SLA, defined per contract
- ✓Dedicated AI infrastructure engineer
- ✓24/7 support · dedicated incident response
All plans include the full audit ledger and signed proofs. Prices in EUR, excl. VAT. Cancel anytime.
Questions before you start
What counts as an AI operation?
Every governed LLM call routed through Azoth — a decision, a block, a cache hit, or a proof. You are billed on what the runtime actually handles, not on idle seats.
Do I have to change my application code?
No. Azoth sits in front of your providers as a drop-in gateway — point your SDK or base URL at it and every call is governed, routed and logged. Your application logic stays exactly as it is.
Can I self-host?
Yes. Enterprise ships as a sovereign runtime you deploy in your own VPC or on-prem — OVH, Scaleway and Outscale compatible — so no governed traffic ever leaves your perimeter.
What happens when I hit a plan limit?
Nothing breaks. Operations keep flowing; you get an in-app prompt to upgrade and a clear view of where the overage is coming from. No surprise throttling mid-incident.