About
Kimi K2 Thinking is Moonshot AI’s most advanced open reasoning model to date, extending the K2 series into agentic, long-horizon reasoning. Built on the trillion-parameter Mixture-of-Experts (MoE) architecture introduced in...
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
OpenAI: o3 ProQ:-0.2
$35.00+3156%
OpenAI: GPT-5 MiniQ:+0.3
$0.69-36%
GLM-5 (Non-reasoning)Q:-0.3
$1.55+44%
Gemini 3 Pro Preview (low)Q:+0.4
$4.50+319%
Performance
132
tokens/sec
Faster than 65% of models
0.86
seconds
Faster than 63% of models
16.05
seconds
Faster than 30% of models
Market Median
94 tok/s
40% faster
Median TTFT
1.11s
23% faster
Throughput/Dollar
122
tok/s per $/1M
Speed Comparison
Molmo2 8B
131 tok/s-0%
IBM: Granite 4.1 8B
132 tok/s+1%
Qwen: Qwen3 Coder Next
133 tok/s+1%
Context Window
262K
tokens
Larger than 62% of models
Max Output
262K
tokens
100% of context
Benchmarks
MMLU-Pro
84.8%
GPQA Diamond
83.8%
HLE
22.3%
LiveCodeBench
85.3%
SciCode
42.4%
TerminalBench Hard
31.1%
MATH-500Not evaluated
AIMENot evaluated
AIME 2025
94.7%
IFBench
68.1%
Long Context Recall
66.3%
Tau2
93.0%
Market AverageTop Score