AI Model Pricing Comparison Misunderstanding | Generated by AI
Question: why xiaomi mimo often cheap than minimax m2.7 though minimax m2.7 show more expensive in million token input and output
Answer: Based on current pricing data and model specifications, there is a misunderstanding regarding which model is more expensive. MiniMax-M2.7 is actually significantly cheaper per million tokens than Xiaomi MiMo-V2-Pro. The discrepancy in “perceived” cost versus “actual” cost likely stems from how each model handles reasoning tokens and their respective pricing tiers on platforms like OpenRouter.
1. Pricing Comparison (Per 1 Million Tokens)
| Model | Input Price | Output Price | Context Window |
|---|---|---|---|
| MiniMax-M2.7 | ~$0.30 | ~$1.20 | 204,800 |
| Xiaomi MiMo-V2-Pro | $1.00 | $3.00 | 1,048,576 |
As shown above, Xiaomi’s MiMo-V2-Pro is roughly 3.3x more expensive for input and 2.5x more expensive for output than MiniMax-M2.7.
2. The Role of Reasoning Tokens
You mentioned that MiniMax might be more expensive due to a /think or reasoning mechanism. Here is how that affects the total cost:
- Reasoning-Heavy Models: MiniMax-M2.7 is a “reasoning-only” or reasoning-optimized model. While its price per token is lower, it generates a much higher volume of hidden internal reasoning tokens (the “thinking” process) before providing a final answer.
- Token Efficiency: Even though MiniMax-M2.7 uses ~55% more output tokens than its predecessor (M2.5) to “think” through a problem, it remains more token-efficient than other reasoning models like GLM-5.
- Xiaomi’s Approach: Xiaomi MiMo-V2-Pro also has a Reasoning Mode, but it typically uses fewer tokens to reach an answer compared to MiniMax. However, because its base price is so much higher ($3.00 vs $1.20), the “cheaper” total for Xiaomi only happens on very short, simple tasks where reasoning is minimal.
3. Summary of Cost Differences
- Why MiniMax seems “more expensive” in practice: If you use MiniMax for a complex task, it might generate 2,000 “thinking” tokens to give you a 100-token answer. You are billed for all 2,100 tokens.
- Why Xiaomi is objectively more expensive: For any fixed amount of data (e.g., a 10,000-word document), Xiaomi will always cost more to process because its base rate per token is higher.
- The Context Factor: Xiaomi offers a much larger 1-million token context window, which requires more expensive hardware to maintain, contributing to its higher $1/$3 price point.
References:
- MiniMax M2.7 - API Pricing & Providers - OpenRouter
-
[Xiaomi: MiMo-V2-Pro – Provider Status and Pricing OpenRouter](https://openrouter.ai/xiaomi/mimo-v2-pro/providers) - MiniMax M2.7 vs MiMo v2 Pro - Pricing & Benchmark Comparison 2026