Models

available models

price per 1M tokens · billed per 1k
MODELPROVIDERINPUT / 1MOUTPUT / 1M
loading…

auto routing

Send model: "agi" and the farm picks the model per request: it classifies the task, filters by capability (vision, context window), enforces the tier policy, then ranks candidates by measured quality, latency and cost. You are billed the same flat rate regardless of which model serves.