Ant Group Digital Technologies has officially launched its decision model LingDT-Decision-1.0, available in 4B and 27B parameter versions on the company's DTMaaS platform. Designed for rapid decision-making tasks, the model shifts from a "generate first, then parse" approach to "directly judging among limited options," reducing token consumption by over 90% and improving response speed by up to 10x in real business tests across customer service, risk control, and finance.
In intelligent customer service scenarios, risk identification accuracy rose from 76% to 89%, while hallucination recall in financial scenarios increased from 44% to 70%. LingDT-Decision-1.0 is the world's first decision model featuring native multi-select and sequential decision-making (DAG) capabilities, enabling it to identify multiple requests simultaneously and complete multi-step decisions based on task dependencies. In evaluations across 23 mainstream open-source benchmarks, the 27B version ranked first in overall score, surpassing Microsoft-Decision-1, Jev 1.13.0, and Fastino/GLiDE models of comparable parameter scale.
Ant Group Releases LingDT-Decision-1.0, Outperforming Microsoft and JEV Models in Decision Accuracy
Aviso legal: El contenido de Phemex News es únicamente informativo.No garantizamos la calidad, precisión ni integridad de la información procedente de artículos de terceros.El contenido de esta página no constituye asesoramiento financiero ni de inversión.Le recomendamos encarecidamente que realice su propia investigación y consulte con un asesor financiero cualificado antes de tomar cualquier decisión de inversión.