Models

Leading models, one integration.

0 models across 0 major families. Switching between them is a one-line change — the same key, the same base URL, the same code.

What it costs

Priced for: An agent run with tool calls — 100,000 input and 8,000 output tokens.

Pick a different shape of work

A typical chat turn One coding task with file context An agent run with tool calls 1M input + 1M output tokens
Pricing is not published yet. We are in a private beta and have not set public prices for these models. When we do, this table will show the dollar cost for the workload you pick alongside each vendor's own published price, so you can check the arithmetic yourself.

The catalog is being finalised

Models appear here once they are qualified and approved for launch.

Request beta access

Access is invite-only while we roll out. This is the catalog planned for launch and may change before general availability. Managed smart routing is an upcoming capability and is not active.