# MiniMax-M2

> MiniMax's text model for agentic work, function calling and advanced reasoning with real-time streaming.

## Key facts

- Model author: MiniMax
- Input: $0.30 per 1M tokens
- Output: $1.20 per 1M tokens
- Cached input: $0.03 per 1M tokens
- Context window: 200,000 tokens
- Maximum output: 128,000 tokens
- Input modalities: text
- Output modalities: text
- Capabilities: reasoning, tools
- Released: 2025-10-23
- Model IDs: `minimax/MiniMax-M2`
- Model page: https://www.realrelay.ai/model/minimax/minimax-m2

## Overview

MiniMax-M2 is a text model from MiniMax whose documentation lists agentic capabilities, function calling, advanced reasoning and real-time streaming. It takes text input and returns text. The 128,000-token output limit includes the model's chain-of-thought, so long reasoning and long answers share one budget.

It has a 200,000-token context window and can generate up to 128,000 output tokens. It accepts text input and returns text. It was released on October 23, 2025.

## Capabilities

MiniMax-M2 supports reasoning and tool calling, so it can plan a task, call the functions you define and use the results in its next step. Real-time streaming lets an application show the answer as it is generated.

Input is text only. Because the output limit counts reasoning tokens as well as the visible answer, it is worth leaving room for the model's thinking when you set output limits.

## Pricing

MiniMax-M2 is priced at $0.30 per 1M input tokens and $1.20 per 1M output tokens. Cached input reads cost $0.03 per 1M tokens.

MiniMax-M2 is available through RealRelay's OpenAI-compatible API.

## Use cases

Use MiniMax-M2 for agent workflows that rely on function calling, such as assistants that search internal data, call business APIs and summarise the results for a user.

Its large output allowance, which includes reasoning, also suits long-form tasks like drafting detailed reports or working through a multi-step analysis in one response.

## Call it

OpenAI-compatible chat completions:

```bash
curl https://www.realrelay.ai/v1/chat/completions \
  -H "Authorization: Bearer $REALRELAY_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model": "minimax/MiniMax-M2", "messages": [{"role": "user", "content": "Hello"}]}'
```

## FAQ

### What is MiniMax-M2?

MiniMax-M2 is a text model from MiniMax whose documentation lists agentic capabilities, function calling, advanced reasoning and real-time streaming. It takes text input and returns text. The 128,000-token output limit includes the model's chain-of-thought, so long reasoning and long answers share one budget.

### How much does MiniMax-M2 cost?

MiniMax-M2 is priced at $0.30 per 1M input tokens and $1.20 per 1M output tokens. Cached input reads cost $0.03 per 1M tokens.

### What are the context window and maximum output of MiniMax-M2?

MiniMax-M2 has a 200,000-token context window and can generate up to 128,000 output tokens.

### What inputs and outputs does MiniMax-M2 support?

MiniMax-M2 accepts text input and returns text.

### Does MiniMax-M2 support function calling?

Yes. MiniMax-M2 supports function calling, reasoning and real-time streaming, with text input and text output.

### When was MiniMax-M2 released?

MiniMax-M2 was released on October 23, 2025.

### How do I call MiniMax-M2 through the API?

Call MiniMax-M2 through RealRelay's OpenAI-compatible API by setting the model field to minimax/MiniMax-M2.