PANews reported on April 30 that Xiaomi's official account "Xiaomi MiMo" has released the first large model MiMo-7B designed specifically for "Reasoning". It surpassed OpenAI o1-mini and Ali Qwen-32B in AIME mathematical reasoning and LiveCodeBench programming assessments. MiMo improves reasoning capabilities by linking large-scale reasoning data pre-training with reinforcement learning post-training, proposes innovative algorithms such as Test Difficulty Driven Reward, and leads in reinforcement learning potential among similar 7B models. The entire series has been open sourced to Hugging Face , and related technical reports have been released simultaneously.