About
Kimi K2 0905 is the September update of [Kimi K2 0711](moonshotai/kimi-k2). It is a large-scale Mixture-of-Experts (MoE) language model developed by Moonshot AI, featuring 1 trillion total parameters with 32...
Related Models
Pricing
Input
$0.60
per 1M tokens
Output
$2.50
per 1M tokens
Blended
$1.07
per 1M tokens
Cheaper than 35% of models. Median price is $0.54/1M tokens.
Cost Calculator
Tokens per day1M
100K100M
Daily
$1.07
Monthly
$32.25
vs. Similar Models
Claude 3.7 Sonnet (Non-reasoning)Q:0.0
$6.00+458%
OpenAI: o1Q:-0.1
$26.25+2342%
Qwen3.5 35B A3B (Non-reasoning)Q:-0.1
$0.69-36%
Claude 4.5 Haiku (Non-reasoning)Q:+0.2
$2.00+86%
Performance
26
tokens/sec
Faster than 1% of models
1.63
seconds
Faster than 31% of models
1.63
seconds
Faster than 55% of models
Market Median
94 tok/s
73% slower
Median TTFT
1.10s
48% slower
Throughput/Dollar
24
tok/s per $/1M
Speed Comparison
Gemma 3 12B Instruct
26 tok/s+1%
MoonshotAI: Kimi K2 0711
26 tok/s+3%
Qwen3.5 2B (Non-reasoning)
27 tok/s+3%
Context Window
262K
tokens
Larger than 62% of models
Max Output
262K
tokens
100% of context
Benchmarks
MMLU-Pro
81.9%
GPQA Diamond
76.7%
HLE
6.3%
LiveCodeBench
61.0%
SciCode
30.7%
TerminalBench Hard
23.5%
MATH-500Not evaluated
AIMENot evaluated
AIME 2025
57.3%
IFBench
41.7%
Long Context Recall
52.3%
Tau2
73.4%
Market AverageTop Score