Ant Group and Renmin University of China jointly released the first native MoE diffusion language model
2025-09-12 10:51:52

On September 11th, at the 2025 Bund Conference, Ant Group and Renmin University of China jointly released "LLaDA-MoE," the industry's first natively MoE-based diffuse language model (dLLM). Experimental data showed that the LLaDA-MoE model outperformed diffuse language models such as LLaDA1.0/1.5 and Dream-7B in coding, mathematics, and agent tasks, approaching or surpassing the autoregressive model Qwen2.5-3B-Instruct. It achieved the performance equivalent to a 3B dense model with only 1.4B parameters activated.
Email Subscription
Newsletters and emails are now available! Delivered on time, every weekday, to keep you up to date with North American business news.
ASIA TECH WIRE

Grasp technology trends

Download