Ant Digital Technology has launched the LingDT-Decision-1.0 model, specifically designed for rapid industrial decision-making tasks. The model is available in two versions, 4B and 27B, and is currently accessible on the company's DTMaaS platform for developers to try for free for a limited time.
The model features native support for multiple-choice and sequential decision-making, which the company claims is better suited for complex enterprise scenarios. In testing across 23 open-source benchmarks, the 27B version of LingDT-Decision-1.0 achieved higher accuracy than other models like Microsoft-Decision-1 and Jev 1.13.
In real-world business applications such as customer service, risk management, and finance, the model demonstrates significant efficiency gains. Compared to traditional large language models, LingDT-Decision-1.0 reduces token consumption by over 90 percent and improves response time by 10 times.
