Write your rules once.

The sidecar is free and open source under MIT. Enterprise adds the control plane: it takes every limit off, delivers your rules to every sidecar you run, and shows you exactly which rules are live and where.

Every rule, every sidecar, one place.

Open Source runs one sidecar with one rule of each kind, from a file on disk. Enterprise takes every limit off, writes the rules for all your sidecars at once, and tells you what each one resolved.

Enterprise

Talk to us

Requires a license

The same sidecar with no rule limits, and a control plane to see and manage every one of them from one place.

  • Unlimited sidecars, guardrails, masking and analyzer rules
  • The full AI Analyzer, including warn and defer
  • Rules written once, delivered everywhere, plus rule sets
  • Every sidecar in one place, with what each one resolved
  • New protocols built for your stack
  • Deployed in your own infrastructure, with support and SLAs
Talk to us

We give you a number for your environment.

Open Source

Free, with limits

MIT, self-hosted

The complete sidecar. Data Masking, Guardrails, AI Analyzer and session audit, running next to your resource.

  • One sidecar, configured from one file
  • One guardrail and one data masking rule
  • One AI Analyzer rule, with the block and allow actions
  • Community support
Get started free

No credit card.

One of each, or no limits.

Open Source gives you one of everything, enough to put hoop in front of one system and see every control work. Enterprise takes the limit off all of them.

What each tier allows
Open SourceEnterprise
Sidecars1Unlimited
Guardrails1Unlimited
Data masking rules1Unlimited
AI Analyzer rules1Unlimited

Take the limits off.

Unlimited sidecars and rules, written once and delivered everywhere, with support behind it. Tell us about your environment and we will give you a number.

Ask an assistant to explain hoop.dev pricing: Claude, ChatGPT, Grok, Gemini, Perplexity, Copilot.

# config.yaml on your free sidecar

# one Data Masking rule
mask:
  enabled: true
  rules:
    - {name: emails, entity: EMAIL_ADDRESS, strategy: redact}

# one Guardrail
policy:
  enforce: true
  rules:
    - name: no-destructive-sql
      type: operation
      operations: [drop, delete, truncate]
      message: destructive statements are not permitted