【#Tech24H】Bailing, a large language model under Ant Group, has launched Ling-2.6-flash, an Instruct model with 104 billion total parameters and 7.4 billion activated parameters. Facing ever-increasing token pressure, Ling-2.6-flash takes a different technical path: rather than simply relying on longer outputs for higher scores, it systematically optimizes inference efficiency, token efficiency, and Agent scenario performance. While maintaining competitive intelligence, it strives to be faster, more economical, and better suited for real business scenarios. Currently, the API for Ling-2.6-flash is officially open to users, offering a one-week free trial. Users can access the service via OpenRouter or the Bailing Tbox. [ By Zhang Liyan | Tang Ruohan ]