Providers & models

Every model, one endpoint,
on ground you control.

Frontier labs, open weights, and your own self-hosted clusters — compared on context, price, and sovereignty, and routed through a single OpenAI-compatible API.

—

providers, one pool

—

routable models

1 API

OpenAI-compatible endpoint

payloads retained

— models across — providers

Model

Context

$/M in

$/M out

Sovereignty

Prices are pass-through provider rates per million tokens. Any OpenAI-compatible endpoint plugs in — add your own model behind a base URL.

Start free

Route to any of them in one request.

Point your SDK at Cudator, name a model or a policy, and we handle the rest — failover, residency, and billing included.

Start free Read the docs

No card to start · multi-currency wallet · cancel anytime

Every model, one endpoint,on ground you control.

Route to any of them in one request.

Every model, one endpoint,
on ground you control.