About
MiMo-V2-Pro is Xiaomi's flagship foundation model, featuring over 1T total parameters and a 1M context length, deeply optimized for agentic scenarios. It is highly adaptable to general agent frameworks like...
Related Models
Pricing
Input
$1.00
per 1M tokens
Output
$3.00
per 1M tokens
Blended
$1.50
per 1M tokens
Cheaper than 29% of models. Median price is $0.54/1M tokens.
Cost Calculator
Tokens per day1M
100K100M
Daily
$1.50
Monthly
$45.00
vs. Similar Models
DeepSeek: DeepSeek V4 FlashQ:0.0
$0.11-93%
Z.ai: GLM 5.1Q:-0.1
$1.50+0%
MiMo-V2.5Q:-0.2
$0.17-88%
OpenAI: GPT-5.2-CodexQ:-0.2
$4.81+221%
Performance
44
tokens/sec
Faster than 11% of models
2.11
seconds
Faster than 22% of models
68.60
seconds
Faster than 3% of models
Market Median
95 tok/s
54% slower
Median TTFT
1.11s
91% slower
Throughput/Dollar
29
tok/s per $/1M
Speed Comparison
Qwen: Qwen3 Max Thinking
44 tok/s+0%
Claude Opus 4.7 (Non-reasoning, High Effort)
44 tok/s-1%
QwQ 32B-Preview
43 tok/s-2%
Context Window
1.0M
tokens
Larger than 90% of models
Max Output
131K
tokens
13% of context
Benchmarks
MMLU-ProNot evaluated
GPQA Diamond
87.0%
HLE
28.3%
LiveCodeBenchNot evaluated
SciCode
42.5%
TerminalBench Hard
40.9%
MATH-500Not evaluated
AIMENot evaluated
AIME 2025Not evaluated
IFBench
68.8%
Long Context Recall
60.7%
Tau2
95.0%
Market AverageTop Score