MiniMax Speech 2.5 Launches: More Expressive Multilingual Capabilities and More Lifelike Voice Cloning

MiniMax has released its next-generation speech synthesis model, Speech 2.5, once again setting a new benchmark for the world’s most powerful voice models. Speech 2.5 achieves three major breakthroughs: enhanced multilingual expressiveness, more accurate voice cloning, and broader language coverage. It supports 40 languages, delivers best-in-class performance in Chinese, significantly improves English similarity, and leads the industry in voice cloning accuracy—capable of preserving accents and speaking styles across languages.