In Hangzhou, China, Ant Group unveiled Ling-3.0-Flash, a new native hybrid-reasoning foundational model for AI agent workflows. This cutting-edge model, designed for quick response times, acts as a high-speed execution node providing a superior blend of intelligence density and cost-effectiveness. With an impressive 124B total parameters, Ling-3.0-Flash operates with just 5.1B active parameters per token, resulting in exceptional performance.