LLM Pricing

← All Models

Xiaomi: MiMo-V2.5-Pro

🇨🇳 Xiaomi · MiMo-V2.5-Pro

Input Price $0.435 per million tokens NT$13.9
Output Price $0.870 per million tokens NT$27.8
Context Window 1.05M tokens Output limit: 131K
OpenRouter Route Price Please verify with official pricing pages
Use this model via OpenRouter →

Overview

Xiaomi: MiMo-V2.5-Pro is a large language model API from Xiaomi, part of its MiMo-V2.5-Pro model family. At $0.435 per million input tokens and $0.870 per million output tokens, it sits in the budget tier — among the cheaper options for high-throughput or cost-sensitive workloads. Output tokens cost about 2× as much as input, so prompt-heavy workloads run noticeably cheaper than generation-heavy ones. An exceptionally large 1.05M-token context window (≈1,573 pages of text) means entire repositories or document collections can be processed without chunking. On Artificial Analysis's Intelligence Index it scores 42 (B grade), a useful proxy for its general reasoning strength relative to the other models tracked here. All prices on this page reflect OpenRouter's routed rates and are re-synced automatically every day; confirm against the provider's official pricing before committing to production.

Dimension Unit Price (USD) Price (TWD) Effective From
Input per 1M tokens $0.435 NT$13.9
Output per 1M tokens $0.870 NT$27.8
Cached Input per 1M tokens $0.0036 NT$0.12

Provider
小米 (Xiaomi)
Model Family
MiMo-V2.5-Pro
Version String
xiaomi/mimo-v2.5-pro
Status
Active
Modality
Text
Context Window
1,048,576 tokens
Output Limit
131,072 tokens

Index Metrics

Cross-domain capability indexes evaluated by Artificial Analysis — Artificial Analysis

Agentic Index 29 C Measured: 2026-07-01
Coding Index 60 S Measured: 2026-07-01
Intelligence Index 42 B Measured: 2026-07-01

Benchmark Scores

Data source: Artificial Analysis

AA-LCR 73.3% B Measured: 2026-07-01
GPQA Diamond 86.6% S Measured: 2026-07-01
HLE 33.8% S Measured: 2026-07-01
IFBench 79.9% A Measured: 2026-07-01
Non-Hallucination 75.4% Measured: 2026-07-01
Omniscience Accuracy 22.6% Measured: 2026-07-01
SciCode 50.2% A Measured: 2026-07-01
Tau2 94.2% Measured: 2026-07-01
TerminalBench 43.2% Measured: 2026-07-01

Performance Metrics

Real-world benchmarks, updated every 72 hours by Artificial Analysis — Artificial Analysis

First Token Latency 2.6s Measured: 2026-07-01
Output Speed 51 t/s Measured: 2026-07-01
Response Time 51.9s Measured: 2026-07-01

Use Case Analysis

Model strengths, weaknesses, and ideal use cases based on benchmarks and official documentation

Strength

  • Ultra-Long Context Reasoning 1M token context window, scores 0.56 BFS/0.92 Parents at 512K. Supports long-horizon tasks with 1000+ tool calls. Official
  • Exceptional Coding Performance SWE-bench Verified 78.9%, GDPVal-AA ELO 1581. Top-tier among open-source models. Official
  • Token Efficiency Uses 40-60% fewer tokens than Claude Opus 4.6, Gemini 3.1 Pro, GPT-5.4 at comparable capability levels. Official
  • Open Source MIT License Fully open-source, MIT license allows commercial integration and weight modification. Official
  • Multimodal Support Native support for text, image, audio, and video input. Official

Best For

  • Complex Software Engineering SWE-bench Verified 78.9%. Ideal for complex long-horizon coding tasks like compiler development, video editors. Official
  • Autonomous Agent Systems Supports long-horizon autonomous tasks with 1000+ tool calls. Has "self-correcting" discipline. Official
  • Cost-Sensitive High-Performance Tasks API pricing $1/$3 per MTok (under 256K), cheaper than comparable closed-source models. Official

Weakness

  • Reasoning Latency Deep thinking mode can result in delays of several minutes before the model begins generating text. Official
  • Hardware Requirements Hosting the 1.02T parameter model locally requires multi-GPU clusters. Official

Not Recommended

  • Real-time Chat Applications Deep thinking mode latency makes it unsuitable for applications requiring immediate responses. Official
  • Budget-Constrained Small Teams Local deployment requires multi-GPU clusters, which is costly. Consider using API or smaller models. Official

90-Day Price Trend

Input / Output price (USD per 1M tokens)

Past 90 days of records; every price change is shown here

Date Dimension Price (USD) Source
Cached Input $0.0036 OpenRouter
Output $0.870 OpenRouter
Input $0.435 OpenRouter
Cached Input $0.0036 OpenRouter
Output $0.870 OpenRouter
Input $0.435 OpenRouter
Cached Input $0.0036 OpenRouter
Output $0.870 OpenRouter
Input $0.435 OpenRouter
Cached Input $0.0036 OpenRouter
Output $0.870 OpenRouter
Input $0.435 OpenRouter
Cached Input $0.0036 OpenRouter
Output $0.870 OpenRouter
Input $0.435 OpenRouter
Cached Input $0.0036 OpenRouter
Output $0.870 OpenRouter
Input $0.435 OpenRouter
Cached Input $0.0036 OpenRouter
Output $0.870 OpenRouter
Input $0.435 OpenRouter
Cached Input $0.0036 OpenRouter
Output $0.870 OpenRouter
Input $0.435 OpenRouter
Cached Input $0.0036 OpenRouter
Output $0.870 OpenRouter
Input $0.435 OpenRouter
Cached Input $0.0036 OpenRouter
Output $0.870 OpenRouter
Input $0.435 OpenRouter
Cached Input $0.0036 OpenRouter
Output $0.870 OpenRouter
Input $0.435 OpenRouter
Cached Input $0.0036 OpenRouter
Output $0.870 OpenRouter
Input $0.435 OpenRouter
Cached Input $0.0036 OpenRouter
Output $0.870 OpenRouter
Input $0.435 OpenRouter
Cached Input $0.0036 OpenRouter
Output $0.870 OpenRouter
Input $0.435 OpenRouter
Cached Input $0.0036 OpenRouter
Output $0.870 OpenRouter
Input $0.435 OpenRouter
Cached Input $0.0036 OpenRouter
Output $0.870 OpenRouter
Input $0.435 OpenRouter
Cached Input $0.0036 OpenRouter
Output $0.870 OpenRouter

Key Insights

Key data points from this page for quick reference and citation.

  • Xiaomi: MiMo-V2.5-Pro Input price: $0.435/M tokens
  • Xiaomi: MiMo-V2.5-Pro Output price: $0.87/M tokens
  • Context window: 1,048,576 tokens
  • Provider: Xiaomi
  • Model family: MiMo-V2.5-Pro
  • Modalities: Text
  • Data source: OpenRouter, updated daily