Skip to content
OmniRail

Every model on the menu.

Text is per million tokens (input / output), images per image at default settings, video per second. Click a model for its full specs, prices and ready-to-copy request.

auto · Not sure which one? Let OmniRail pick per request, at the chosen model's price.How it chooses →

New category

Decision models

Not chat models. You send what is going on and a few typed questions; you get back numbers your code can branch on: the probability that a statement is true, which of your options fits, or where it sits on your scale. No prose to parse.

Billed per input token, the answers are free. Same key, same balance.

New category

Uncensored models · fewer refusals

Open models tuned to follow your instructions instead of hedging: villains that stay villains, critique without the cushioning, straight answers on health, law or history. Ordinary chat models: same key, same API and MCP tool, paid per call.

These models refuse less often, but our usage policy still applies to every call: nothing illegal, nothing sexual involving minors, no harassment or threats against real people, no malware or weapons meant to hurt anyone. Usage policy

Loading the catalog…