— PRICING · THREE PATHS

No checkout. Three paths.

Every paid engagement starts with a 30-minute call so the node is configured against your threat model — not a marketing page.

— ENGAGEMENT PATHS

Start free. Scale when ready.

Path 01

Audit

Free · No account

Paste a system prompt or live HTTPS endpoint. RAE attacks it with 50–200 adversarial prompts across 10 categories. Severity-ranked findings report, async.

  • NO ACCOUNT
  • NO CREDIT CARD
  • WORK EMAIL REQUIRED
  • 5–10 MIN TURNAROUND
Run a free audit →
Path 02Primary

Node

Per-agent · Annual

On-prem Docker node between your application and LLM provider. Sub-50ms p95. Five detector heads. Four actions.

  • OPENAI-COMPATIBLE PROXY
  • TS + PYTHON SDK
  • SIDECAR MODE
  • 5 DETECTOR HEADS
  • 4 ACTIONS
  • 30-MIN ONBOARDING CALL
Book a demo →
Path 03

Enterprise

Volume · Air-gapped · Custom

Kubernetes Helm chart, air-gapped mode for regulated environments, custom detector heads trained on your taxonomy.

  • KUBERNETES HELM CHART
  • AIR-GAPPED MODE
  • CUSTOM DETECTORS
  • GOVERNMENT READY
  • VOLUME PRICING
Contact sales →

FIG. 01 — engagement ledger · no self-serve checkout at launch

— COMMON QUESTIONS

Common
questions.

A 30-minute call to configure the node against your threat model. After that: change your base URL (OpenAI-compatible proxy) or add the SDK. No other code changes.

Per active agent, annual contract. Volume discounts for fleets of 5+ agents. Contact us with your expected traffic and we'll scope it properly.

RAE operates entirely within your network boundary — no outbound connections to the RAE control plane. Metadata never leaves. Required for certain government and regulated-industry deployments.

Yes. The audit report is the first thing we discuss on the onboarding call. Most Node deployments start with an audit.

— GET STARTED

Start with a free audit.

Run a free auditBook a demo