Architecture pattern
Cascade router
A cheap decision chooses whether to use rules, a small model, a premium model, a specialist, or a human.
Plausible
What the pattern does
A cheap decision chooses whether to use rules, a small model, a premium model, a specialist, or a human.
Caveat
Bad routing can erase all cost savings through failures and retries.
Patterns are recurring shapes observed across the collected material. A repeated shape is evidence that developers converge on it, not evidence that it performs well in production.
Sources
Limitations
- Bad routing can erase all cost savings through failures and retries.
Where this pattern appears
- system-one-router ObservedA published npm extension (v0.4.5) for the Pi coding agent that adds a per-turn model router; its README states it is based on and retains attribution to an existing router, yeliu84/pi-model-router.
- Agent Model Router Authored HypothesisA drop-in router that selects model, reasoning effort, and fallback policy per step.
- The strongest near-term architecture is Jev as a complement and control layer around generative models. PlausibleJev produces decisions rather than prose; Vercel and OpenRouter expose it through evaluation-oriented endpoints, and LangChain documents it as a routing and guardrail layer inside its own harness.
- jev-router ObservedAn npm-published CLI wrapper that launches Claude Code or OpenAI Codex through a loopback proxy, with a documented test suite covering shared policy, both request formats, and decision display.