modelbenchmark.io

Kimi K2 0905

Kimi · kimi-k2-instruct

Compare

Kimi model for long-context chat, coding, and agentic reasoning

Specification

most-agreed values

Context
262K
Max output
100K
Released
2025-09-25
Knowledge cutoff
2024-10
Retires
2026-04-15
Open weights
yes
Input
text
Output
text

Price

US dollars per million tokens · most-agreed

Input
$0.40
Output
$1.80
Cache read
$0.20
Cache write
$0.78

Available from 14 hosts

HostIn $/MOut $/MCache rdCache wrContextOutputRetires
nvidia · moonshotaifreefree262K262K
io-net · moonshotai$0.39$1.90$0.195$0.7833K4K
nano-gpt · moonshotai$0.40$1.80$0.20
262K
256K
128K
100K
8K
deepinfra$0.50$2.00$0.40
131K
262K
131K
262K
jiekou · moonshotai$0.57$2.30131K131K
novita$0.57$2.30131K100K
novita-ai · moonshotai$0.57$2.30131K33K
baseten$0.60$2.50
fireworks_ai$0.60$2.50
131K
262K
16K
33K
wandb · wandb$0.60$2.50128K128K
groq$1.00$3.00$0.50262K16K2026-04-15
huggingface · moonshotai$1.00$3.00
262K
131K
16K
together_ai$1.00$3.00262K2026-03-06
hyperbolic$2.00$2.00131K131K

More from Kimi

most-hosted first