It can both "answer multiple-choice questions" and "answer fill-in-the-blank questions" — Xingchen Intent Understanding Model Xing4.0-29B-Intent released

What happened: China Telecom AI's Xingchen AGI Lab released and open sourced Xing4.0-29B-Intent, a dedicated intent understanding model that combines intent recognition with parameter extraction and outputs structured JSON. Key numbers: 2.75 million training corpus entries; 6,000 questions from 48 public datasets in the first evaluation; 5,606 Why it matters: The lab says dedicated small models are faster, cheaper and more reliable than general models for converged, latency-sensitive decision tasks, and the model is already used in telecom customer service and smart speaker products.

China Telecom AI's Xingchen AGI Lab released and open sourced Xing4.0-29B-Intent, a dedicated intent understanding model that combines intent recognition with parameter extraction and outputs structured JSON.

The lab says dedicated small models are faster, cheaper and more reliable than general models for converged, latency-sensitive decision tasks, and the model is already used in telecom customer service and smart speaker products.