+THE ROUTING STUDIO

See the decision.
Follow the route.

A clear view of every decision. Set a policy, follow the selected path, and compare three sample providers.

Routing workbenchSAMPLE
YOUR RULES. ONE ENDPOINT.
ROUTING POLICY
Choose what matters.

Weigh estimated cost, latency, and reliability together.

PROVIDER AVAILABILITY

Switch it off to test an unavailable route.

Sample routing.
No inference sent.
SELECTED ROUTEBalanced

Trusted peer

Best combined sample score: 40% estimated cost, 40% latency, and 20% reputation.

Your toollocal-chat
cairn
Local modelEligible
01
Your APIEligible
02
Trusted peerSelected
03
3 eligible destinationsSample request path

Provider comparison

REQUEST ESTIMATES
Sample provider estimates, ranked by the active policy. Reliability is a simple reputation proxy.
ProviderEst. costLatencyReliability*
Trusted peerSelected
Est. cost$0.0010Latency410 msReliability*98.8%
Configured APIEligible
Est. cost$0.0032Latency240 msReliability*99.5%
Your local modelEligible
Est. cost$0.0000Latency840 msReliability*96%

*Sample reliability is a reputation proxy. Bars compare sample latency against 840 ms.

Inspect decision
Model alias
local-chat
Policy
Balanced
Outcome
Sample route selected
  • Trusted peerBest combined sample score: 40% estimated cost, 40% latency, and 20% reputation.
  • Configured APIEligible under this policy.
  • Your local modelEligible under this policy.

This preview uses illustrative scoring. It does not contact your router or send an inference request.

Illustrative estimates. No models or API keys required. This preview demonstrates policy tradeoffs, not production scoring.
+TRY THIS

Close the cloud door.

Choose Local only. Remote candidates become ineligible. Then switch your local model offline: the request stops instead of falling back to a remote provider.

+READ THE NUMBERS

Estimates, with context.

Prices are synthetic request estimates. Latency is sample time to response headers. The demo uses reliability as a simple stand-in for reputation.

+A DELIBERATE CHOICE

One alias. Your substitutes.

These providers share the sample alias local-chat. In Cairn, you choose which models can substitute for one another. A shared alias does not promise equal quality.

How does this demo choose a route?

Balanced uses an illustrative normalized score: 40% estimated cost, 40% latency, and 20% reputation. Lowest cost or Lowest latency ranks eligible providers by that single criterion. Local only filters out remote providers before selection.

All numbers and identities are sample fixtures. This page makes no provider calls and does not connect to your local router. The actual router uses configured prices, estimated token usage, operator priors, and locally observed reliability. The peer price illustrates a future paid-capacity scenario; public-chain payments remain disabled.

+YOUR MACHINE. YOUR DIRECTION.

Ready to find your own route?

Start with the synthetic beta demo, then connect your own models.

Beta setup