Xiaomi Unveils Three MiMo-V2 Large Models; Lei Jun Announces $16B Boost for AI

Xiaomi’s ambitious large-model plans were fully revealed in the spring of 2026.
On March 19, Xiaomi officially launched three self-developed large models: MiMo-V2-Pro, MiMo-V2-Omni, and MiMo-V2-TTS. This is more than a routine tech upgrade — it marks a milestone in Xiaomi’s deep commitment to the “Agent era.”
To underline its AI focus, Xiaomi’s founder stated on social media that the company’s R&D and capital investment in AI this year will exceed 16 billion yuan. He also noted that the trillion-parameter MiMo-V2-Pro ranks eighth globally in Artificial Analysis’ comprehensive intelligence benchmark and fifth in brand ranking worldwide.
Each of the three models plays a distinct role, forming a complete Agent stack:
Flagship base model MiMo-V2-Pro: Designed for high-intensity Agent tasks. With over 1 trillion (1T) total parameters, it uses a hybrid attention mechanism, balancing efficiency and capacity with 42B activated parameters. It supports an ultra-long context of 1 million Tokens, excelling in complex logical reasoning and tool calling.
Omni-modal base model MiMo-V2-Omni: Natively integrates text, vision, and audio. It spans the entire pipeline from sensory understanding to action execution, making it essential for Agents to perceive the physical world.
Speech large model MiMo-V2-TTS: Gives Agents “warm” expression abilities with fine-grained emotional control, creating more human-like machine interactions.
On the commercialization front, Xiaomi has adopted an extremely aggressive pricing strategy. Input within a 256K context costs only $1 per million Tokens, significantly undercutting competitors at the same tier. Both the Pro and Omni versions have officially opened API services.
Notably, the driving force behind this elite AI team is known as the “AI prodigy.” The mysterious model “Hunter Alpha,” which stirred the developer community, is actually the internal test version of MiMo-V2-Pro.
Related article
DeepMind CEO Hassabis: I sleep six hours a day, usually feel energetic around 1 a.m.
Fortune recently featured an interview with Demis Hassabis, CEO of Google DeepMind, revealing his unconventional approach to rest and productivity. Hassabis disclosed that he sleeps very little, structuring his waking hours into two distinct work blo
OpenAI, Anthropic Vie for Market Share Despite Revenue Shortfalls
Despite recent reports suggesting OpenAI missed revenue targets, creating pressure on tech stocks this Tuesday, private AI lab investors remain resilient. Seasoned backers have confirmed they will not reduce investment despite negative media coverage
California AV Compliance: A New Era of Tickets, Geofences, and 1M Miles
Guident operates an AuveTech shuttle in South Florida, managing a four-mile route in West Palm Beach and a one-mile route in Boca Raton using its remote monitoring technology. | Credit: GuidentCalifornia is redefining the regulatory landscape for dri
Related Special Topic Recommendations
Comments (0)
0/500

Xiaomi’s ambitious large-model plans were fully revealed in the spring of 2026.
On March 19, Xiaomi officially launched three self-developed large models: MiMo-V2-Pro, MiMo-V2-Omni, and MiMo-V2-TTS. This is more than a routine tech upgrade — it marks a milestone in Xiaomi’s deep commitment to the “Agent era.”
To underline its AI focus, Xiaomi’s founder stated on social media that the company’s R&D and capital investment in AI this year will exceed 16 billion yuan. He also noted that the trillion-parameter MiMo-V2-Pro ranks eighth globally in Artificial Analysis’ comprehensive intelligence benchmark and fifth in brand ranking worldwide.
Each of the three models plays a distinct role, forming a complete Agent stack:
Flagship base model MiMo-V2-Pro: Designed for high-intensity Agent tasks. With over 1 trillion (1T) total parameters, it uses a hybrid attention mechanism, balancing efficiency and capacity with 42B activated parameters. It supports an ultra-long context of 1 million Tokens, excelling in complex logical reasoning and tool calling.
Omni-modal base model MiMo-V2-Omni: Natively integrates text, vision, and audio. It spans the entire pipeline from sensory understanding to action execution, making it essential for Agents to perceive the physical world.
Speech large model MiMo-V2-TTS: Gives Agents “warm” expression abilities with fine-grained emotional control, creating more human-like machine interactions.
On the commercialization front, Xiaomi has adopted an extremely aggressive pricing strategy. Input within a 256K context costs only $1 per million Tokens, significantly undercutting competitors at the same tier. Both the Pro and Omni versions have officially opened API services.
Notably, the driving force behind this elite AI team is known as the “AI prodigy.” The mysterious model “Hunter Alpha,” which stirred the developer community, is actually the internal test version of MiMo-V2-Pro.
DeepMind CEO Hassabis: I sleep six hours a day, usually feel energetic around 1 a.m.
Fortune recently featured an interview with Demis Hassabis, CEO of Google DeepMind, revealing his unconventional approach to rest and productivity. Hassabis disclosed that he sleeps very little, structuring his waking hours into two distinct work blo
OpenAI, Anthropic Vie for Market Share Despite Revenue Shortfalls
Despite recent reports suggesting OpenAI missed revenue targets, creating pressure on tech stocks this Tuesday, private AI lab investors remain resilient. Seasoned backers have confirmed they will not reduce investment despite negative media coverage





Home






