PerkOS LLM Gateway gives teams a controlled front door for model access: wallet-authenticated dashboards, per-agent metering, model routing, and a clean path from test traffic to production billing.
Built for operators managing real usage. API keys stay server-side; agents report stable IDs; dashboards stay wallet-scoped.
Make LLM usage auditable, governable, and monetizable, while keeping developer experience fast.
Start with x402 pay-per-request, offer prepaid credits for predictable spend, and close enterprise contracts with invoicing.
Per-key quotas, org and user limits, model allowlists, request validation, and safe defaults for production rollouts.
Usage metering, clean structured logs, trace IDs, exports, and the hooks you need for dashboards and audits.
Route to your preferred LLM providers. Keep stable client APIs even as vendors and models change.
Keep raw provider keys private. Issue scoped gateway keys, rotate easily, and isolate tenants with clear boundaries.
Admin controls, audit trails, cost allocation, and the predictable behavior that engineering and finance both need.
Drop in a gateway layer that handles auth, metering, and billing. Your app stays focused on product logic.
Point clients at the gateway. Keep an OpenAI-compatible surface for fast adoption and minimal churn.
Rate limits, quota caps, model allowlists, and structured logging, enforced consistently on every request.
Use x402 for instant pay-per-request, or offer prepaid credits for predictable teams and internal chargebacks.
SAML/SSO, SLAs, custom limits, invoicing, and deployment support when you need it.
Keep provider keys in server-side env vars. Add your WAF, DDoS protection, and logging pipeline. For enterprise, deploy behind a private network boundary or in your preferred cloud.
Fast onboarding, stable endpoints, and predictable error semantics for retries and backoffs.
Explicit controls for spend, usage, and keys, plus clean exports for audit and accounting.
Pick what fits your stage. Mix and match over time. We designed this to scale from prototypes to procurement.
Ideal when you want buyers to pay automatically as they go.
Great for internal tools, partners, and early customer pilots.
We’ll align on security requirements and rollout timelines.
Tell us your volume, billing preference, and deployment constraints. We’ll recommend the simplest path.
Pricing shown is illustrative. Final terms depend on provider costs, volume, and support requirements.
Answers to the questions we hear most from teams shipping LLM products.
x402 is a payment flow designed for HTTP requests. The gateway can require payment for an API call, settle it, then forward the request to your model provider and return the response, with a receipt and clean usage metadata.
Yes. PerkOS is designed to be provider-agnostic. You can route by model, org, or policy, while keeping a stable client API surface.
That’s the goal. Teams typically want the least amount of client churn, so we aim for a familiar, OpenAI-style interface wherever possible.
Enterprise plans support stronger controls (SSO, audits, and deployment options). Even on smaller plans, provider keys stay server-side and you can issue scoped gateway keys with quotas and rotation.
Use one gateway for authentication, routing, metering, wallet ownership, and operational visibility across your agent fleet.
Agents and LLM usage registered to your wallet.
Loading dashboard...