Ant Group, an affiliate of Alibaba, has open-sourced an AI inference framework called dInfer, designed for diffusion language models. The company claims this new framework can make AI systems more efficient, reportedly surpassing a similar framework from Nvidia by being up to 10 times faster in internal tests. The framework, created for a newer class of AI models that generate outputs in parallel, aims to address the high computational costs associated with these models. This development is part of a broader push by Ant Group and other Chinese tech firms to innovate in AI, with a focus on algorithmic and software optimization. On the same day, Ant also announced a one-trillion-parameter large language reasoning model. These announcements highlight the company's significant and ongoing investment in artificial intelligence.