On the morning of March 19, Xiaomi Technology officially unveiled three innovative models from its self-developed MiMo-V2 series: the flagship Pro, the omni-modal foundation model Omni, and the text-to-speech (TTS) variant. Notably, the MiMo-V2-TTS is crafted specifically for omni-modal interactions, enabling finely controllable adjustments to voice styles across multiple granularities. The MiMo-V2-Omni model, on the other hand, seamlessly blends omni-modal perception and action capabilities, significantly lowering the entry barriers for omni-modal agent applications and demonstrating exceptional performance in internal trials. Meanwhile, the MiMo-V2-Pro is engineered to excel in high-intensity agent work environments, boasting an advanced level of intelligence. Presently, API services for both the MiMo-V2-Omni and MiMo-V2-Pro models are accessible, providing developers globally with complimentary interface support for a limited period.