Priced on spend governed, not seats.
Flat monthly tiers that scale with the agent spend we attribute and cap for you — never a markup on the tokens themselves, never per-seat math. Start free.
Free
See what every agent spends. Free forever.
Start free- Up to $500/mo of governed agent spend
- Meter every call via the SDK or gateway
- Spend attributed per agent, model & customer
- Budgets in alert mode
- Bring your own provider keys — no token markup
- Signed run traces · 1 connector · 7-day history
Team
For teams that need control, not just visibility.
Upgrade to Team- Up to $5,000/mo of governed agent spend
- Budgets that block — runaway calls stopped before they run
- Delegation grants & mesh audit: who called whom, on whose behalf
- Chargeback-ready spend exports
- Team workspaces: members, roles, projects
- Unlimited private agents & connectors · 90-day history
- Priority support
Scale
For fleets whose agent spend is a real line item.
Upgrade to Scale- Up to $50,000/mo of governed agent spend
- Everything in Team
- 1-year run & spend history
- Priority support with faster SLAs
Enterprise
For orgs with compliance, scale, and audit demands.
Contact sales- Unlimited governed agent spend
- Everything in Scale, plus —
- SSO / SAML & SCIM provisioning
- Org-wide audit logs & granular RBAC
- VPC or self-hosted deployment
- SLA & dedicated support
Compare every plan
Questions, answered.
What counts as “governed agent spend”?
The provider dollars (OpenAI, Anthropic, Google, …) that flow through Silvanexum’s metering — what we attribute, budget, and audit each month. You pay the providers directly at their rates; our tiers scale with how much spend we govern for you, not with seats. Limits are soft during beta: if you outgrow a tier we’ll talk, nothing gets cut off mid-month.
Do you mark up model tokens?
No. You bring your own provider keys, so you pay OpenAI, Anthropic, or Google directly at their rates. Silvanexum charges for the platform — spend attribution, traces, governance, collaboration — never for the tokens your agents spend.
Is the free tier really free?
Yes, free forever. You can meter your agents through the SDK or gateway, see spend attributed per agent and per model, and set alerting budgets without entering a card.
What happens when an agent hits its budget?
Your choice, per budget: alert (you're notified, the agent keeps running) or block (the call is rejected before it executes, so the spend never happens). Blocked calls return a clear budget_exceeded error your code can handle gracefully.
What happens to my provider keys?
Keys are encrypted with AES-256-GCM, validated on paste with a free probe, and decrypted only server-side at the moment an agent runs. They're never exposed to the browser and never sent anywhere but the provider you chose.
Can I self-host Silvanexum?
Enterprise plans can deploy into your own VPC or fully self-hosted, with SSO, audit logs, and an SLA. Talk to us about your environment.
Bring your keys. We'll bring the spend control.
Free to start, no card required. Upgrade when your team needs budgets that block, delegation audit, and shared workspaces.