Continual Platform

Models

Available models

One model. Use this id in Continual Platform requests. Rates are per 1M tokens and separate cache-miss input from cheaper cache reads.

Priced at cost. We serve the model ourselves and charge what it costs us to run your request — no margin on top. Rates are per 1M tokens; every request's real cost is on the Usage page.

ModelTypeInput miss / 1MCache read / 1MOutput / 1M
mgpt-1-27bText$0.425$0.085$2.55