MiniMax LLM¶

The minimax_llm module integrates MiniMax models into AutoRAG via the OpenAI-compatible API.

Supported Models¶

Model	Context Window
MiniMax-M2.7	1,048,576 tokens
MiniMax-M2.7-highspeed	1,048,576 tokens
MiniMax-M2.5	1,048,576 tokens
MiniMax-M2.5-highspeed	204,800 tokens

Features¶

Auto-truncate prompt¶

Prompts that exceed the model’s token limit are automatically truncated to prevent API errors.

Temperature clamping¶

MiniMax models accept temperature values between 0 and 1. Values above 1.0 are automatically clamped to 1.0.

Think-tag stripping¶

MiniMax M2.5+ models may include <think>...</think> reasoning tags in their output. These are automatically stripped from the generated text.

Module Parameters¶

llm: The MiniMax model name. For example, MiniMax-M2.7 or MiniMax-M2.5-highspeed.
batch: The batch size for API calls. Default is 16.
truncate: Whether to truncate input prompts to the model’s max length. Default is True.
api_key: MiniMax API key. You can also set this to env variable MINIMAX_API_KEY.
And all parameters from the OpenAI Chat Completion API (MiniMax uses an OpenAI-compatible endpoint).

Example config.yaml¶

modules:
  - module_type: minimax_llm
    llm: [MiniMax-M2.7, MiniMax-M2.5-highspeed]
    temperature: [0.1, 0.5]
    max_tokens: 512
    api_key: ${MINIMAX_API_KEY}