Pricing
Priced like a pentest. Delivered like a platform.
Starting points, not quotes. Every engagement is scoped first, and the scope sets the price. Nothing below is a subscription to a tool. It is a human red team with a dashboard attached.
One engagement
Rapid Agent Pentest
From $3,500
One AI agent, one clear surface. Pre-launch or post-update.
- Up to 2 agent endpoints or flows
- Security and judgement scenarios
- Roughly 1 week from kickoff to readout
- Retest window: 30 days
One engagement
Comprehensive Red Team
From $6,500
An agent with tools, money, or customer data behind it. Multiple flows, multiple languages.
- Up to 5 agent endpoints or flows
- Security, judgement, and business impact assessment
- Dual-solve on high-stakes scenarios
- Roughly 2 weeks from kickoff to readout
- Retest window: 60 days
Retainer
Continuous
From $1,500 / month
You ship agent changes often and want every release tested against the attacks that already worked.
- Monthly scenario refresh against your live agent
- Retest of every closed finding on every release
- Standing dashboard and regression suite
- Priority queue for new findings
Included in every tier
Retesting is not an add-on. A finding you cannot verify as fixed is not a finding you can act on.
- Human red team, no SDK or integration required
- Every P0 and P1 finding reproduced by a lead auditor before you see it
- Live findings dashboard with resilience score
- Reproduction steps and a recommended fix on every finding
- Free retest of every fix, finding closes only when the attack no longer lands
- Regression suite export (JSON and JSONL) for your CI pipeline
- Webhooks for your own tooling
Why publish prices?
Because you should know whether this is a conversation worth having before you have it.
What moves the price?
Number of agent flows, tools the agent can call, languages in scope, and whether high-stakes scenarios need two independent testers.
Early partners
We are onboarding a small number of early partners at reduced rates in exchange for a reference. Ask.