GPT-5.6 Luna is the fastest and most cost-efficient model in the GPT-5.6 family. It delivers strong capability for everyday, high-volume workloads while keeping latency and cost low.