Pricing
Agent Authorization
as a Service
Secure every AI agent decision with identity, delegation, trust scoring, risk evaluation, and audit evidence.
No infrastructure. No policy engine to manage.
Your agents call our API — we handle the rest.
Start with a sandbox account
Get full access to every AGF feature to evaluate the platform — no credit card, no commitment. After 30 days, upgrade to go live.
- 300 authorization decisions / month
- 2 active agents
- 30-day trial window
- All enterprise features enabled
- agfk_test_ key prefix
- Audit artifacts flagged non-compliance
Sandbox → live in one admin toggle
Starter
For teams validating agent governance.
Decisions / month
2,000
Agents
5
Included
- Agent identity & verified passports
- Delegation token issuance
- Trust scoring per session
- Three-layer risk evaluation
- Signed, replayable audit trail
- Audit log export (JSON · CSV)
- Email support
Growth
Most PopularFor production multi-agent systems.
Decisions / month
15,000
Agents
50
Included
- Everything in Starter
- Stop compromised agents instantly
- Track every delegated action
- Git-backed policy versioning
- Agent registry with role assignments
- Configurable delegation chains
- Rego policy authoring
- Priority support
Enterprise
For regulated industries.
Decisions / month
Unlimited
Agents
∞
Included
- Everything in Growth
- Prove every decision to auditors
- NIST AI RMF · ISO 42001 · EU AI Act
- Multi-organization agent federation
- Policy drift detection & alerting
- Automated incident runbooks
- Dedicated SLA & support
- Private deployment option
All prices in USD. Decisions reset monthly. Need more? Contact us for volume pricing or annual discounts.
Compare plans
| Feature | Starter | Growth | Enterprise |
|---|---|---|---|
| Authorization decisions / month | 2,000 | 15,000 | Unlimited |
| Agents | Up to 5 | Up to 50 | Unlimited |
| Agent identity & passports | |||
| Delegation tokens | |||
| Trust score computation | |||
| Three-layer risk evaluation | |||
| Signed audit artifacts | |||
| PDP REST API | |||
| Audit log export (JSON · CSV) | |||
| Stop compromised agents instantly | |||
| Configurable delegation chains | |||
| Rego policy authoring | |||
| Git-backed policy versioning | |||
| Real-time anomaly alerting | |||
| Compliance reporting (NIST · ISO · EU) | |||
| Multi-organization federation | |||
| Policy drift detection | |||
| Dedicated SLA & support | |||
| Private deployment |
Why it matters
The cost of ungoverned agents
Without AGF
- No agent identity — you can't tell which agent made a decision
- No delegation controls — sub-agents inherit unlimited authority
- No audit evidence — you can't prove what happened in a review
- No trust model — all agents treated equally regardless of behaviour
With AGF
- Verified agent identities with cryptographic passports
- Scoped delegation with configurable depth limits
- Signed, replayable decision artifacts for every action
- Risk-aware authorization that adapts to agent behaviour
FAQ
What is a decision?
Every call to POST /v1/decide counts as one authorization decision — whether the result is allow or deny. Decisions reset at the start of each billing month.
Do I need to run infrastructure?
No. AGF is fully managed — no policy engine, no database, no servers to operate. Your agents call our API and we handle authorization, revocation, trust scoring, and audit storage.
Can I self-host?
Private deployment is available on Enterprise plans only. Contact us to discuss your requirements.
Do you support compliance frameworks?
Yes. Enterprise includes compliance reporting mapped to NIST AI RMF, ISO 42001, and the EU AI Act — including audit evidence packages suitable for formal regulatory review.
Ready to govern your agents?
We're onboarding pilot partners for the managed Starter tier now. Enterprise conversations welcome.