Pricing

Enterprise protection.
No enterprise price tag to start.

Start with 1 million AI requests protected — completely free. No credit card. Full features. No capability limits.

Free
$0
1M requests or 30 days — the full product, no limits on capability
  • Prompt injection detection (2-tier ML + regex)
  • Deep DLP — NER + pattern matching + secret scan
  • Content safety classification (5 categories)
  • Response filtering
  • OPA / Rego policy enforcement
  • ZKP audit trail & evidence bundles
  • OpenAI & Anthropic native adapters
  • Real-time threat dashboard
  • Community support
Start Free — No Card

No credit card required. Full feature access.

Most Popular
Startup
Flat rate/mo
Predictable pricing for AI-native teams past their free tier
  • Everything in Free
  • Unlimited protected requests
  • Multi-environment deployment (dev/staging/prod)
  • Advanced analytics dashboard
  • Custom Rego policy templates
  • Priority email support (48h SLA)
  • Webhook alerting for blocked threats
  • MCP server protection
  • Compliance export (GDPR, HIPAA reports)
Get Started

Contact us for exact pricing.

Enterprise
Custom
Full governance platform. SLAs. Self-hosted. Dedicated support.
  • Everything in Startup
  • Self-hosted / data residency deployment
  • SSO / SAML integration
  • Dedicated SLA (99.9% uptime guarantee)
  • Compl-AI compliance automation
  • Remed-AI automated remediation
  • Trace-AI software supply chain
  • Dedicated security engineer onboarding
  • Custom contract and procurement
Contact Sales

Includes access to the full Zerberus platform.

Full feature comparison.

Feature Free Startup Enterprise
Detection & Protection
Prompt injection detection ✓ Full ✓ Full ✓ Full
ML classification (Prompt Guard 2) ✓ ✓ ✓
Deep DLP — NER + pattern matching ✓ Full ✓ Full ✓ Full
Secret scanning (API keys, tokens) ✓ ✓ ✓
Content safety classification ✓ 5 categories ✓ 5 categories ✓ Custom categories
Response filtering ✓ ✓ ✓
MCP server protection ~ Basic ✓ Full ✓ Full
Policy & Governance
OPA / Rego policy enforcement ✓ Templates ✓ Custom ✓ Full
Multi-tenant policy isolation — ✓ ✓
Rate limiting by tenant/user ✓ ✓ ✓ Custom
Policy version control + rollback — ✓ ✓
Compliance & Audit
ZKP-based evidence bundles ✓ ✓ ✓
Immutable audit logs ✓ 30-day retention ✓ 1-year ✓ Custom retention
GDPR compliance exports — ✓ ✓
HIPAA compliance exports — ✓ ✓
EU AI Act documentation support — ~ Partial ✓ Full
Infrastructure
Managed proxy deployment ✓ ✓ ✓
Self-hosted / on-premise — — ✓
Data residency (EU, US) — ~ US only ✓ Custom
SLA uptime guarantee — ~ 99.5% ✓ 99.9%
SSO / SAML — — ✓
Compl-AI + Remed-AI + Trace-AI — — ✓ Full platform

Frequently asked questions.

What counts as a request?
A request is one API call to an LLM provider routed through RAGuard — typically one user message or agent action. Streaming responses count as one request regardless of token count. Responses (the LLM's reply) are processed as part of the same request and don't count separately.
Does the free tier have any feature limits — or just volume limits?
Volume limits only. The free tier includes full access to all RAGuard capabilities: prompt injection detection, deep DLP, content safety, OPA policy enforcement, response filtering, and ZKP audit trails. No capability degradation.
How does RAGuard affect my LLM API costs?
RAGuard doesn't add tokens to your prompts or responses — it inspects them inline without modification (unless redaction is triggered by your policy). Blocked requests don't reach your LLM provider, which can reduce costs when injection attacks are prevented. Forwarded requests are billed by your provider as normal.
Can I use RAGuard with models other than OpenAI and Anthropic?
RAGuard currently has native adapters for OpenAI and Anthropic APIs. Support for additional providers (Gemini, Mistral, open-source models via Ollama and vLLM) is on the roadmap. Contact us if you have a specific provider requirement.
What's the difference between RAGuard and Portkey or other AI platforms?
Portkey and similar tools are optimised for routing, reliability, observability, and AI infrastructure operations. RAGuard is designed for runtime governance: governing prompts, context, tool access, policy decisions, and governance evidence in production. They solve different layers and can be complementary. See our comparison pages →
Is there a self-hosted option for data residency requirements?
Yes — available on the Enterprise plan. RAGuard can be deployed within your own AWS, Azure, or GCP environment, ensuring AI interaction data never leaves your infrastructure. Container images and deployment documentation are provided. Contact sales for details.
What compliance frameworks does the audit trail support?
RAGuard's ZKP-based audit trail is designed to support GDPR Article 22 (automated decision-making documentation), HIPAA Technical Safeguard requirements, and EU AI Act Article 12 (transparency logging for high-risk AI). Formatted compliance exports are available on Startup and Enterprise tiers.

Still have questions?

Talk to the team. We'll walk you through the right configuration for your stack, answer security questionnaires, and help you set up your free trial.

Start Free Now Talk to Sales