Home Product Compliance Pricing Blog Research
Pricing

Start free. Scale as you grow.

Early access customers are onboarded personally. Paid tiers add higher volumes, the dashboard and compliance features as they ship. Indicative early pricing, shown to illustrate the model.

Early pricing, subject to change at launch
Developer
£0/month
Free forever
  • 10,000 agent calls/month
  • Core injection detection
  • Basic audit logging
  • Compliance reports
  • MCP governance
Startup
£99/mo
Billed monthly
  • 100,000 agent calls/month
  • 4 live detection layers
  • Tool call policy engine
  • Monthly compliance report (roadmap)
  • MCP governance (roadmap)
Team
£599/mo
Billed monthly
  • Unlimited agent calls
  • Agreed support SLA
  • Slack support channel
  • Custom policies + audit assist
  • Dedicated onboarding
FAQ

Questions a security
team asks first.

Straight answers to the things that matter before you route a single agent call through us.

Where is my data stored, and in what region?
If you self host, everything stays on your own infrastructure, in the region you choose. For hosted deployments we agree the region with you during onboarding. The audit log records metadata only: the agent, the tool, the decision, the layer that made it, the reason and a timestamp, with each entry hash chained to the one before. Prompt and output content is not written to the log.
Does Reinward see the content of my prompts?
The gateway inspects content in memory at the moment of the call, which is how injection and PII detection work, but that content is not persisted. Your audit trail records the decision and its metadata, never your prompt text, and the SHA 256 hash chain makes any later tampering with the log detectable.
What's the real latency impact?
The live layers are rule based and run inside the gateway with no external calls, so the overhead is small next to the multi second round trip of an LLM agent call. An optional semantic scanner, off by default, adds an external model call, so it stays off unless you enable it. We will publish measured latency figures from our pilot deployments.
Can I self-host or run it in my own VPC?
Yes. Reinward runs on your own infrastructure, including inside your own VPC, so agent traffic never leaves your network perimeter. Today we deploy it with you during onboarding. A packaged, self serve install is on the roadmap.
What happens if Reinward goes down? Does my agent stop?
You decide in your integration. If the gateway cannot be reached, your code can block the tool call (fail closed, recommended for agents with destructive tools) or let it through and record the gap (fail open) for non critical workloads.
Which frameworks does it actually work with?
Anything that can make an HTTP request before running a tool: LangChain, LlamaIndex, CrewAI, AutoGen, n8n and custom agents. Because Reinward sits outside the framework rather than hooking into its internals, you are not locked into one ecosystem. A Python SDK is on the roadmap.

Your agents are running.
Your audit trail isn't.

Early access open. No credit card. Guided onboarding.

UK-built | Designed for UK GDPR | Maps to SOC 2 and EU AI Act controls

You're on the list, we'll be in touch soon.