MiniMax M1
Neural Network
MiniMax M1 is an open reasoning model from MiniMax for long context tasks, available via MiniMax M1 API.
Max answer length
(in tokens)
Context size
(in tokens)
Prompt cost
(per 1M tokens)
Answer cost
(per 1M tokens)
How it works MiniMax M1?
MiniMax M1 is an open-weights reasoning model from the MiniMax team, built on a hybrid Mixture-of-Experts architecture with lightning attention. This combination provides cost-effective inference on long inputs: the model is designed for extended context and step-by-step task analysis. Weights are published, making it easy to compare answers with closed models on your own data. In BotHub, you work with MiniMax M1 without VPN or foreign cards, pay with Russian cards in rubles per token, and unused tokens do not expire. All models are in one window, switching takes a few clicks, and a unified OpenAI-compatible API allows changing models in a multi-model system without rewriting integration. For teams, we offer contracts, invoices, electronic document management, an admin panel with limits, and AES-GCM encrypted chats. Useful when you need to analyze a long document, build a chain of reasoning in analytics, verify technical logic, or build your own assistant via API.Frequently asked questions about MiniMax M1
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.
minimax-m1 is a text model with reasoning and a context window of up to a million tokens. It is suitable for working with large documents and codebases, refactoring and debugging code, STEM tasks, analytics, and agent scenarios with tool calling. It also accepts images and documents as input.
Up to 40,000 tokens in a single response — enough for long technical analysis, detailed documentation, a large code module, or an extensive analytical report. If the material exceeds this volume, break the task into parts: the million-token context allows you to keep the entire conversation history.
Yes, in our data, the model has a reasoning mode enabled: it builds a chain of steps before the final answer. This significantly helps in mathematics, logic, requirements analysis, and code debugging, where it is important to break down the task rather than providing the first option that comes to mind.
Yes. The model supports function calling and structured JSON output, making it easy to integrate into agents, chatbots, and pipelines where the response goes directly to a database or service. In BotHub, this works via a unified OpenAI-compatible API.
According to our data, the model accepts text, images, and documents as input: you can send an interface screenshot, diagram, chart, or a large file with requirements and ask it to analyze the content. For audio and video, use specialized models in BotHub — switching takes a couple of seconds.
The model is multilingual and works normally with Russian queries: formulate your task in your usual language. It is easy to check the quality on your own texts — in BotHub, you pay per token rather than for a subscription, and you can compare the answer with another model in the same window.