Everything that moves a request is free on every plan — routing, failover, live swap, your own keys at zero markup. The subscription pays for the layer that judges, compares, and proves your AI works.
The plumbing. Route production AI traffic across every model.
The intelligence. Proof your AI works, on your real traffic.
Pro for the whole org — shared control, shared visibility.
For teams with procurement, security, and scale needs.
| Free | Pro | Team | Enterprise | |
|---|---|---|---|---|
| Included on every plan | ||||
| Gates and projects | Unlimited | Unlimited | Unlimited | Unlimited |
| Every model, live model swap | ||||
| Failover and spend caps | ||||
| Bring your own keys — zero markup | ||||
| Your own models as routing targets | ||||
| Claude Code, Cursor, Codex, MCP connectors | ||||
| Observability — logs, analytics, tracing | ||||
| Usage | ||||
| Units per month | 10M | 250M | 1B | Custom |
| Log history | 7 days | 30 days | 90 days | Custom |
| Seats | 1 | Unlimited | Unlimited | Unlimited |
| Intelligence — Pro and up | ||||
| Quality scoring on real traffic | — | |||
| Experiments and shadow comparisons | — | |||
| Independence Reports | — | |||
| Team — and up | ||||
| Shared gate configs | — | — | ||
| Role-based access control | — | — | ||
| Team spend budgets | — | — | ||
| Enterprise | ||||
| SSO / SAML | — | — | — | |
| SLAs | — | — | — | |
| Self-host options | — | — | — | |
| Security review + DPA | — | — | — | |
| Support and billing | ||||
| Support | Priority email | Priority email | Dedicated | |
| Billing | — | Card, monthly | Card, monthly | Invoice, annual |
Abuse-prevention rate limits apply on every plan — they scale with your tier and are not something we sell. Units never gate routing: on Free they meter tracked traffic, on paid plans they size intelligence coverage.
Open-weight models in your own cloud account. We write the deployment, run it with you, monitor it, upgrade models, and keep parity proven — compute bills to your cloud directly, never through us. From $1,000 setup, then $499/month per environment.
Honestly: this only makes financial sense above roughly $2,000/month of model spend, or when data residency outweighs cost. Below that, serverless open-weights setup (Together, Fireworks, and friends) is included free on every plan — it takes five minutes and we'll do it with you.
Really. When you bring your own provider keys, traffic bills straight to your provider accounts at provider prices — we never touch the invoice. Trial traffic on our keys passes through at provider list price against your $5 credit. The subscription pays for the intelligence layer, not a per-token cut.
A token processed through the gateway — input and output, on any model, including requests routed to your own endpoints. Non-token work like image generation and audio converts at published equivalents in the docs. Units never gate routing: on Free they meter how much of your traffic we track, and on paid plans they size how much traffic the intelligence layer covers. For scale: 250M units is roughly what a Sonnet-based product with a few thousand monthly users burns. Provider cache reads, failed requests, and our own background verification never count against you.
Your traffic is never blocked, on any plan. On paid plans we notify you at 80% and 100%, keep serving and recording at full quality, and never auto-charge an overage — only two consecutive months well over your allowance moves you up a tier, at the next period, with notice. On Free, routing never stops: past 10M units your requests keep flowing and stay in your history, but tracking is locked from view until you upgrade. Downgrades are never automatic — dropping your plan is one click in settings, anytime.
Units already capture how much your team uses the platform — charging per seat on top of that would double-count the same growth. Invite everyone.
Anytime, from Settings → Billing → Manage subscription. You keep your plan until the end of the period you paid for, then land back on Free. Your gates, keys, and analytics are untouched.
Works with every major provider, and 100+ more LLMs.