About
MiniMax-M2.1 is a lightweight, state-of-the-art large language model optimized for coding, agentic workflows, and modern application development. With only 10 billion activated parameters, it delivers a major jump in real-world...
Related Models
Pricing
Input
$0.29
per 1M tokens
Output
$0.95
per 1M tokens
Blended
$0.45
per 1M tokens
Cheaper than 53% of models. Median price is $0.54/1M tokens.
Cost Calculator
Tokens per day1M
100K100M
Daily
$0.45
Monthly
$13.65
vs. Similar Models
DeepSeek V4 Pro (Non-reasoning)Q:-0.2
$0.54+20%
GPT-5 (low)Q:-0.2
$3.44+656%
MiMo-V2-Flash (Reasoning)Q:-0.2
$0.15-67%
Qwen3.6 35B A3BQ:+0.2
$0.35-22%
Performance
225
tokens/sec
Faster than 92% of models
7.70
seconds
Faster than 14% of models
16.59
seconds
Faster than 29% of models
Market Median
94 tok/s
138% faster
Median TTFT
1.11s
592% slower
Throughput/Dollar
494
tok/s per $/1M
Speed Comparison
Nova 2.0 Lite (Non-reasoning)
224 tok/s-1%
Qwen3 0.6B (Reasoning)
224 tok/s-1%
OpenAI: o3 Mini High
223 tok/s-1%
Context Window
205K
tokens
Larger than 57% of models
Max Output
197K
tokens
96% of context
Benchmarks
MMLU-Pro
87.5%
GPQA Diamond
83.0%
HLE
22.2%
LiveCodeBench
81.0%
SciCode
40.7%
TerminalBench Hard
28.8%
MATH-500Not evaluated
AIMENot evaluated
AIME 2025
82.7%
IFBench
69.9%
Long Context Recall
59.0%
Tau2
85.4%
Market AverageTop Score