MiniMax M2 Her
Neural Network
MiniMax M2 Her is a conversational model from MiniMax for roleplay and multi-turn conversations with a consistent character.
Max answer length
(in tokens)
Context size
(in tokens)
Prompt cost
(per 1M tokens)
Answer cost
(per 1M tokens)
How it works MiniMax M2 Her?
MiniMax M2 Her is a language model from MiniMax built around dialogue: roleplay, character chat, and long multi-turn conversations where maintaining character and tone is crucial. The model focuses on expressive lines and rich messages, making it suitable for scenarios requiring a lively conversational style rather than dry facts. In BotHub, it is accessible without a VPN or foreign card: pay with Russian cards in rubles based on usage; tokens do not expire. With over 250 models in one window, you can compare responses and switch models without rewriting your integration: access is via a unified OpenAI-compatible API, and chats are encrypted with AES-GCM. For teams, we offer contracts, invoices, EDI, and an admin panel with limits. It is useful for authors of interactive stories and text games, developers of character-based chatbots, creators of virtual companions and digital characters for apps, screenwriters needing to flesh out character dialogue, and teams defining their brand voice.Frequently asked questions about MiniMax M2 Her
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-m2-her accepts text, documents, and images as input and responds with text: it analyzes materials, helps with code and refactoring, and prepares summaries, explanations, and drafts. The context window is 32,768 tokens, so a long document or a large project fragment will fit into the request.
The single response limit is 2048 tokens. This is enough for a detailed explanation, a significant code snippet, or a medium-length text. If you need longer material, break the task into parts and continue generation with subsequent requests, passing the previous result into the context.
We have no data indicating a separate reasoning mode for this model, but that does not rule out the possibility. In practice, a simple trick helps: ask it to explain the solution step-by-step. This makes answers to logic and STEM tasks clearer and much easier to verify.
There is no information in our data regarding function calling or structured JSON output — the data is simply missing. The fastest way to check is with a couple of requests via the unified BotHub OpenAI-compatible API: if the format doesn't suit you, switch to another model without rewriting your integration.
You can upload images and documents: the model analyzes their content and responds with text. Audio and video input is not noted in our data. Image, audio, or video generation is also not attributed to this model — you always get text as output.
We have no confirmed data on the quality of Russian for this specific MiniMax model, so we will not make any promises. The most reliable way is to run your real request and compare the result with other models in one BotHub window — switching takes a couple of seconds.