The right model for every call. Automatically.
Router sits in the Caveman Platform request path and sends each call to the cheapest model in your pool that passes your evals. If nothing cheaper passes, nothing moves. Router and the Platform are in private development.
Every call goes to your most expensive model. Switching is a risk nobody wants to own.
Most teams pin one model for the whole application, because it is the one they have watched work. A cheaper model would handle a large share of the traffic, but proving that call by call means building the harness, running the comparison, and owning the regression if the answer turns out to be wrong.
So the pin stays, and every call pays the top rate — the hard ones and the trivial ones alike.
The cheapest model that passes your evals. Or traffic stays put.
- the rule
- Router sends each call to the cheapest model in your pool that passes your evals. If nothing cheaper passes, traffic stays put.
- eval-gated rollout
- Record → replay → shadow → canary → active. Each stage has a gate, and a route only reaches live traffic after it clears the stage before it.
- automatic rollback
- A regression rolls the route back without waiting for anyone. The route you were serving keeps serving.
- what we will claim
- Savings stay inferred until they are provider-causal. No invented percentage, and no dollar figure your own traffic has not produced.
- 01recordreal traffic, captured
- 02replaythe candidate, on what already happened
- 03shadowalongside production, serving nothing
- 04canarya slice of live traffic
- 05activethe route, in full
One gate decides. The pool it opens onto is yours.
your traffic, call by call
your model pool, cheapest first
What powers it.
Stop paying the top rate
for easy calls.
Router and Caveman Platform are in private development. Spend is priced from provider-reported usage against the public model catalog; unknown models stay unpriced, and verified savings start at $0.00 and move only on provider-causal evidence.