Pay less than you save
Four plans, one platform. Self-hosted available on Business and Enterprise. Every plan includes the MCP Server, REST API, LiteLLM integration, and x402.
- Up to $1K/mo tracked
- 1 provider
- 5 agents
- 7-day data retention
- 1 alert rule
- Basic cache analysis
- LiteLLM integration
- x402: pay-per-call only
- Up to $10K/mo tracked
- 3 providers
- 25 agents
- 30-day retention
- 10 alert rules
- Advanced recommendations
- Cascade routing insights
- Full cache analysis
- LiteLLM integration
- x402: 10 premium calls/mo
- 5 team members
- Up to $100K/mo tracked
- Unlimited providers
- 100 agents
- 90-day retention
- Unlimited alerts
- Auto-optimization
- Cascade auto-apply
- Cache + auto-suggest
- Self-hosted (Docker)
- x402: unlimited + wallet
- Agent spend policies
- 25 team members
- API access
- Unlimited everything
- 1-year retention
- SSO (SAML)
- Custom integrations
- Self-hosted + custom infra
- Cascade custom rules
- Team-wide cache strategy
- x402: wallet + custom policies
- Outbound buyer mode
- Dedicated support
- SLA guarantee
- 5% auto-savings fee
x402 agent payments — AI agents pay for premium tools via HTTP 402 with USDC. No accounts needed.
Refer other teams and earn credits toward your plan. 10 referrals = 3 months Pro free.
Frequently asked questions
What counts as tracked spend?+
Any AI/LLM API cost that flows through your connected providers (OpenAI, Anthropic, Google AI, etc). We aggregate usage data from their billing APIs.
Can I change plans anytime?+
Yes. Upgrade or downgrade instantly. When downgrading, you keep access until the end of your billing period.
What happens if I exceed my plan limits?+
We never cut off your monitoring. You get a notification to upgrade. Data older than your retention period is archived, not deleted.
Do you store my API keys?+
API keys are encrypted with AES-256 and stored in Supabase Vault. We use read-only access to pull usage data. We never make API calls on your behalf.
How does the MCP Server work?+
Your AI agents connect to our MCP endpoint and can check budgets, report usage, and get recommendations in real-time. Standard MCP protocol — works with Claude, GPT, Gemini.