# MiMo V2.5

1T total, 42B active (MoE). Pro-level agentic performance at roughly a third of MiMo V2.5-Pro's cost.

## Overview

- **Provider:** Xiaomi
- **Model ID:** `mimo-v2.5`
- **Type:** Chat
- **Canonical URL:** https://modelstack.cc/models/chat/mimo-v2.5
- **Version:** V2.5
- **Best for:** Agentic reasoning, step-by-step thinking, tool use, balanced cost-performance

## Pricing

| Usage | Price |
| --- | ---: |
| Input | $0.14 / 1M tokens |
| Output | $0.28 / 1M tokens |
| Cache read | $0.0028 / 1M tokens |
| Cache write (5 minutes) | $0.14 / 1M tokens |

## Specifications

- **Context window:** 1M
- **Maximum output:** 16K
- **Typical latency:** 16 ms
- **Released:** 2026-02

## Capabilities

- chat
- tool_use
- reasoning
- coding
- analysis

## API usage

Use this model ID in a ModelStack API request:

```json
{
  "model": "mimo-v2.5"
}
```

For the rendered page, visit https://modelstack.cc/models/chat/mimo-v2.5.