Xiaomi: MiMo-V2.5-Pro

MiMo-V2.5-Pro is a text-only model from Xiaomi with a context window of roughly one million tokens and a maximum output of 131,000 tokens per response. It supports tool use and reasoning, which makes it suited for multi-step or agentic workflows, though structured output support is unconfirmed. There is no public information on whether weights are open. Benchmark coverage is limited to three benchmarks, so treat the blended score of 74.0 as a partial signal rather than a settled verdict. The coding sub-score of 99.3 stands out and may appeal to developers, but the agentic score of 48.0 suggests caution for complex autonomous tasks. At $0.435 per million input tokens and $0.87 per million output tokens, the price is competitive, making it worth shortlisting for coding-heavy workloads where cost efficiency matters and users are willing to account for thin benchmark coverage elsewhere.