MiMo V2.6 Pro UltraSpeed
Neural Network
MiMo V2.6 Pro UltraSpeed is a fast edition of Xiaomi's flagship: same quality as MiMo V2.6 Pro, with significantly faster generation.
Specifications
- Input
- text, images, documents
- Output
- text
- Context size
- 1 048 576 tokens
- Max answer length
- 131 072 tokens
- Reasoning effort
- low, medium, high
- Capabilities
- reasoning, function calling
Prompt cost
(per 1M tokens)
Answer cost
(per 1M tokens)
How it works MiMo V2.6 Pro UltraSpeed?
MiMo V2.6 Pro UltraSpeed is a high-speed edition of Xiaomi's flagship model, built from the same trillion-parameter checkpoint as MiMo V2.6 Pro. It matches the base version in response quality but delivers results about ten times faster, so the difference lies in response time, not content. Through BotHub, the model is accessible from Russia without a VPN or foreign card: you pay with ruble cards for actual token usage, and they do not expire. Connection is via a unified OpenAI-compatible API, allowing you to switch to any of over 250 models in one window without rewriting integration. Chats are encrypted via AES-GCM, and for companies, we offer contracts, invoices, EDI, and an admin panel with limits. Choose UltraSpeed where every second counts: real-time conversational assistants and support, streaming processing of large text arrays, drafting and editing, code analysis and generation in the editor, and prototyping where you run dozens of prompt variations in a row without wanting to wait.Frequently asked questions about MiMo V2.6 Pro UltraSpeed
You can use generated results for commercial purposes. You own all rights to the content you create. The only restriction: make sure your prompt does not include copyrighted third-party material. You are responsible for respecting the rights to any input data.
This is a Xiaomi text model with reasoning and tool calling. It accepts text, documents, and images as input, and a context of up to 1,048,576 tokens allows for analyzing large files, codebases, and long conversations in their entirety. Suitable for analytics, coding, STEM tasks, and agent scenarios via a unified API.
In a single response, the model outputs up to 131,072 tokens — that's dozens of pages of coherent text. Enough for a large technical document, detailed analysis, or a major code module without splitting it into parts. If the response hits the limit, just ask it to continue from where it stopped.
Yes, the reasoning mode before answering is confirmed in our data. The model first builds a chain of steps and only then formulates the result — this significantly helps with math, logic, code debugging, and tasks where you need to synthesize conditions from a large document.
Yes, function calling and structured JSON output are supported. You describe the schema, and the model returns a predictable object that goes directly into your logic. Through the BotHub OpenAI-compatible API, such an agent connects without rewriting integration, and you can switch models with a single line.
Images and documents — yes: send a screenshot, diagram, chart, or PDF and ask to analyze the content along with a text question. Audio and video input is not noted in our data, so for voiceovers or video analysis in BotHub, it is more convenient to use a specialized model.
We have no data on the quality of Russian language performance, so we won't make any promises — test it on your own tasks. In BotHub, it's fast: access is open from Russia without a VPN or foreign card, payment is in rubles based on usage, and there are other models in the same window for comparison.