Baseline seeded — waiting for the first market change.
GPU market syncing — no offers yet. Admin → Bootstrap GPU market.
Decision Layer

TCO Calculator — Should I switch, or self-host?

Independent, non-conflicted. Real prices from every host. No inference proxied — just the math.

Your workload

State encoded in URL — send it to your team.

All models — cheapest hosts at your volume

Enter volume to compute
Enter a monthly volume to run.
Data updates hourly. WhatModel doesn't proxy inference — this is decision support, not a router.