Xiaomi: MiMo-V2.5

MiMo-V2.5 is a Xiaomi model that accepts text, audio, image, and video input, making it one of the broader multimodal options for teams that need to process mixed media through a single endpoint. It offers a context window of about 1.05 million tokens and allows up to 945,000 completion tokens, so long documents, transcripts, or extended outputs fit without heavy chunking. It supports tool calling and reasoning; structured output support is not specified, and open-weights status is not stated. At $0.119 per million input tokens and $0.238 per million output tokens, it sits at the low end of paid pricing, which makes it worth shortlisting for high-volume multimodal or long-context workloads where cost per call matters. The tradeoff is evidence: it has no independent benchmark coverage yet, so its quality is unproven. Run your own evaluations on representative tasks before committing production traffic.