About
Kimi K2.5 is Moonshot AI's native multimodal model, delivering state-of-the-art visual coding capability and a self-directed agent swarm paradigm. Built on Kimi K2 with continued pretraining over approximately 15T mixed...
Related Models
Pricing
Input
$0.38
per 1M tokens
Output
$2.02
per 1M tokens
Blended
$0.79
per 1M tokens
Cheaper than 43% of models. Median price is $0.54/1M tokens.
Cost Calculator
Tokens per day1M
100K100M
Daily
$0.79
Monthly
$23.63
vs. Similar Models
MiniMax: MiniMax M2.7Q:0.0
$0.32-60%
Z.ai: GLM 5 TurboQ:0.0
$1.90+141%
OpenAI: GPT-5.4 NanoQ:+0.1
$0.46-41%
GPT-5.2 (medium)Q:-0.1
$4.81+511%
Performance
42
tokens/sec
Faster than 9% of models
1.26
seconds
Faster than 41% of models
72.34
seconds
Faster than 2% of models
Market Median
94 tok/s
56% slower
Median TTFT
1.11s
13% slower
Throughput/Dollar
53
tok/s per $/1M
Speed Comparison
OpenAI: GPT-4
42 tok/s+1%
Claude Opus 4.7 (Non-reasoning, High Effort)
42 tok/s+1%
DeepSeek R1 Distill Qwen 32B
43 tok/s+3%
Context Window
262K
tokens
Larger than 62% of models
Max Output
256K
tokens
98% of context
Benchmarks
MMLU-ProNot evaluated
GPQA Diamond
87.9%
HLE
29.4%
LiveCodeBenchNot evaluated
SciCode
49.0%
TerminalBench Hard
34.8%
MATH-500Not evaluated
AIMENot evaluated
AIME 2025Not evaluated
IFBench
70.2%
Long Context Recall
65.3%
Tau2
95.9%
Market AverageTop Score