This is the first open-weight model that plays like a top-class agentic model. Watching it go through ambiguous and meticulous chained tool work successfully puts it squarely in the wheelhouse of Opus 4.6. Weβre looking at an open weight model, but with much cheaper direct inference provider pricing. For a subclass of our eval set, itβs outperforming GPT 5.2. Weβre about to undergo a gigantic industry shift.
Join the Conversation
Share your thoughts and go deeper down the rabbit hole