Xiaomi unveils its MiMo-V2 AI models, including the 1T-parameter MiMo-V2-Pro, codenamed Hunter Alpha, which Xiaomi says benchmarks close to GPT-5.2 and Opus 4.6 (Carl Franzen/VentureBeat)
Why it matters: New AI models from Xiaomi and MiniMax intensify the race for advanced, self-optimizing large language models.
- Xiaomi launched its MiMo-V2 AI models, featuring the 1T-parameter MiMo-V2-Pro (Hunter Alpha), which they claim rivals GPT-5.2 and Opus 4.6 in benchmarks (VentureBeat).
- Fuli Luo, a key figure from the DeepSeek R1 project, is leading Xiaomi's AI initiative, underscoring the strategic importance of this release (VentureBeat).
- MiniMax also released its M2.7, a proprietary "self-evolving" LLM, which it used to build, monitor, and optimize its own reinforcement learning harnesses, showcasing a distinct approach to AI development (TechMeme via VentureBeat).
Xiaomi has unveiled its MiMo-V2 AI models, including the formidable 1T-parameter MiMo-V2-Pro, codenamed Hunter Alpha, which the company claims benchmarks comparably to advanced models like GPT-5.2 and Opus 4.6. This release, led by DeepSeek R1 veteran Fuli Luo, signals a significant push into the competitive AI landscape, while MiniMax also launched its "self-evolving" M2.7 LLM, highlighting a broader trend of innovative AI development.
