Documentation
Kamiwaza Documentation
Everything you need to route your first request, configure routing policies, and connect private GPU infrastructure.
Documentation sections
Quickstart 5 min
Install the CLI, register an endpoint, write your first routing policy, and send your first routed request.
Get started →API Reference
Full reference for all /v1/route, /v1/endpoints, /v1/policies, and /v1/audit endpoints.
Browse API →Routing Policies
YAML policy language reference. Match on data class, tenant ID, latency budget, cost cap. First-match evaluation.
Read policy docs →Private GPU Setup
Connect on-prem Llama or other self-hosted models via VPC peering or PrivateLink.
Setup guide →Tenant Isolation
Configure per-tenant model allowlists, rate limits, and dedicated audit buckets in a shared gateway.
Isolation guide →