Xiaomi: MiMo-V2.6-Pro-UltraSpeed

MiMo-V2.6-Pro-UltraSpeed is a text generation model from Xiaomi, listed on api14 as xiaomi/mimo-v2.6-pro-ultraspeed. It is the fast speed edition of Xiaomi's flagship foundation model, MiMo-V2.6-Pro, and is built from the same 1T MiMo-V2.6-Pro checkpoint. According to Xiaomi, it matches the original model in quality while delivering roughly 10x the speed. It sits in the text category and is aimed at developers who want flagship-level output without the latency of the full model.

The model accepts text, image, video and audio as input, and produces text as output. That makes it a multimodal model on the input side: you can send a prompt that includes an image, a video or an audio clip alongside text, and receive a text response. The context window is 1,048,576 tokens, which is large enough to hold very long documents, long transcripts or extended multi-turn conversations in a single request. The maximum output is 131,072 tokens, so responses can be long as well. Supported request parameters include temperature, top_p, frequency_penalty, presence_penalty, max_tokens, stop, response_format, structured_outputs, tools, tool_choice, reasoning and include_reasoning.

Those capabilities point to a few practical uses. The large context window suits tasks that involve reading and reasoning over long inputs, such as summarising lengthy reports, analysing long transcripts, or answering questions across a large body of text. Because images, video and audio can be passed in, the model can be used for tasks like describing or extracting information from visual media, working with recorded audio, or combining a document with a screenshot in one prompt. Tool calling and structured outputs make it a reasonable fit for agent-style workflows and for applications that need responses in a fixed schema, such as data extraction pipelines. The reasoning parameter gives you a way to request reasoning output where a task benefits from it.

You call the model through api14's OpenAI-compatible API using the model id xiaomi/mimo-v2.6-pro-ultraspeed. Because the endpoint follows the OpenAI chat completions format, you can use existing OpenAI SDKs and point them at api14, or send requests directly. The request body accepts the parameters listed above: temperature, top_p, frequency_penalty, presence_penalty, max_tokens, stop, response_format, structured_outputs, tools, tool_choice, reasoning and include_reasoning. Use max_tokens to cap output length up to the 131,072 token maximum, and use tools with tool_choice when you want the model to call functions you define.

One thing to keep in mind is that the model outputs text only, so it is not the right choice if you need generated images, audio or video. Its input side is multimodal, but the response is always text. If your workload does not need the speed edition, or if you need output modalities other than text, another model may fit better. For text tasks that involve long inputs, mixed media input, tool use or structured output, this model covers those cases directly.