The 1.5T parameter Grok 4.6 model launched today with a specific focus on long-running agents. It targets the knowledge work sector, claiming higher efficiency than its primary competitors. xAI optimized the architecture for ambitious interactive tasks. This release positions the model as a direct challenger to the current top-tier coding and reasoning agents.