Simple, transparent pricing
The runtime is open source and free forever. Pay only when you want the platform watching over it.
Free
The open-source runtime
- Open-source CLI & runtime - yours forever
- Command interception & policy enforcement
- Secret masking & cloaking
- Prompt injection detection
- Tamper-evident audit trail, stored locally
Everything runs on your machine. No account needed, no data leaves your laptop.
Starter
The platform, for one person
First month free - then $15/month for 1 user and up to 5 agents.
- Everything in the open-source runtime
- Live dashboard - watch every agent & tool call
- MCP gateway - one policy layer in front of every MCP server
- MCP Hub - see every server & tool, deny any of them from the console
- Contextual security - rules that understand your sessions & repos
- Skill checker - scans agent skills for hidden risks
- Session history with 30-day retention
Fair limits
- 1 user
- Up to 5 agents
- 100,000 tool-interception events / month
- 1,000 skill scans / month
Enterprise
For teams & fleets
- Everything in Starter - with no usage caps
- Managed service across every user, device & agent in your fleet
- Contextual learning - allow/deny rules that sharpen as usage compounds
- Per-session taint control over tool calls, tracked independently
- Business context drift detection, tuned to your organization
- Per-policy human-in-the-loop approval
- Open telemetry routing - plug agent events into any SIEM
- Prismor-hosted evaluation - guard agents in any language with no sidecar
- Teams, SSO & SCIM with org-wide policy hierarchy
- Hardened rules that agents cannot disable
- Priority support & on-prem deployment
Pricing FAQ
What do I get for free?
The full Prismor runtime is open source: command interception, policy enforcement, secret masking, prompt-injection detection, and a local audit trail. It protects your AI coding agents on your machine, forever, with no account required.
What does Starter add on top of open source?
Starter connects your agents to the Prismor platform: a live dashboard of every tool call, the MCP gateway and MCP Hub so every MCP server your agents reach sits behind one policy layer, contextual security that understands your sessions and repos, skill scanning, and 30 days of hosted history. It covers one user with up to 5 agents, 100,000 tool-interception events, and 1,000 skill scans per month. The first month is free, then $15/month.
How does the free first month work?
Every new workspace gets its first month of Starter free. After that it's $15/month for 1 user and up to 5 agents, and you can drop back to the open-source tier any time.
What happens if I hit a Starter limit?
Your agents stay protected - the open-source runtime keeps enforcing policy locally no matter what. The dashboard warns you as you approach a monthly cap; past it, new events pause until the next cycle or you can talk to us about Enterprise, which has no caps.
Can I cancel any time?
Yes, from the billing portal in your dashboard. Your platform access continues to the end of the paid period, then your workspace returns to the free open-source tier. Local protection never stops.