modelbenchmark.io

Kimi K2.5 (TensorX)

Kimi · kimi-k2-5

Compare

Earlier Kimi frontier model for long-context agents, coding, and multimodal work

Specification

moonshot listing

Context
262K
Max output
262K
Released
2026-01-27
Knowledge cutoff
2025-01
Retires
2026-08-31
Open weights
yes
Input
text, image
Output
text

Price

US dollars per million tokens · moonshot list price

Input
$0.60
Output
$3.00
Cache read
$0.10
Cache write
free

Quality

6 benchmarks · 2 sources

Composite
75th
percentile of 344 scored models
Rank
87 / 344
±0.05 sd
Evidence
6 × 2
3 or more benchmarks, from 2 or more sources
Effort range
3.0
points between effort settings
coding72nd1/3 bench
reasoning90th1/4 bench
math71st3/7 bench
knowledge38th1/1 bench

Every score

one row per source and configuration — nothing averaged away

BenchmarkScoreConfigurationSourceRun
OTIS Mock AIME 2024-202592.2 ±3.9defaultEpoch AI2026-02-02
GPQA diamond87.6 ±1.9defaultEpoch AI2026-02-02
SWE-Bench verified73.8 ±2.0defaultEpoch AI2026-02-17
↳70.8high · mini-SWE-agentSWE-bench—
Sources and settings disagree by 3.0 points. Both figures stand.
FrontierMath-v127.9 ±2.6defaultEpoch AI2026-02-03
FrontierMath-Tier-4-2025-07-014.2 ±2.9defaultEpoch AI2026-02-02
SimpleQA Verified34.3 ±1.5defaultEpoch AI2026-08-27

The leading figure for a benchmark is the median across its configurations, so one heroic high-effort run cannot set the number.

Available from 48 hosts

HostIn $/MOut $/MCache rdCache wrContextOutputRetires
moonshotvendor$0.60$3.00$0.10—262K262K2026-08-31
alibaba-coding-planfreefreefreefree262K33K—
alibaba-coding-plan-cnfreefreefreefree262K33K—
alibaba-token-planfreefreefreefree262K98K—
alibaba-token-plan-cnfreefreefreefree262K98K—
scnet-token-planfreefreefree—262K262K—
tencent-coding-planfreefreefreefree262K33K—
nano-gpt · moonshotai$0.375$2.00$0.07—262K66K—
nano-gpt · thinking$0.375$2.00$0.07—262K66K—
auriko$0.45$2.25$0.07—262K262K—
deepinfra$0.45$2.25$0.07—262K262K—
deepinfra · moonshotai$0.45$2.25$0.07—262K33K—
edenai$0.45$2.25$0.07—262K262K—
meganova · moonshotai$0.45$2.25$0.07—262K262K—
siliconflow · moonshotai$0.45$2.25$0.07—262K262K—
siliconflow-cn · pro$0.45$2.25$0.07—262K262K—
edenai · tensorx$0.50$2.80$0.125—262K262K—
openrouter · moonshotai$0.50$2.50$0.15—262K236K—
tensorx · moonshotai$0.50$2.80$0.125$0.625262K262K—
venice$0.56$3.50$0.22—256K66K—
alibaba-cn$0.574$3.011$0.1148$0.7175262K33K—
llmgateway$0.574$3.011$0.225—262K262K—
llmgateway-providers
$0.574
$0.60
$3.011
$3.00
$0.10—262K
98K
33K
—
qiniu-ai · moonshotai$0.5797$3.0435——256K256K—
abacus$0.60$3.00——262K33K—
aihubmix$0.60$3.00$0.10—262K33K—
alibaba-cn · kimi$0.60$3.00$0.10—262K262K—
azure$0.60$3.00——262K262K—
azure_ai$0.60$3.00$0.10—262K262K2027-01-26
azure-cognitive-services$0.60$3.00——262K262K—
baseten$0.60$3.00————2026-10-16
fireworks_ai$0.60$3.00$0.10—262K33K—
frogbot$0.60$3.00$0.10—256K128K—
hpc-ai · moonshotai$0.60$3.00$0.10—256K256K—
huggingface · moonshotai$0.60$3.00$0.10—262K262K—
jalapeno$0.60$3.00$0.10—262K180K—
jiekou · moonshotai$0.60$3.00$0.10—262K262K—
kilo · moonshotai$0.60$3.00——262K236K—
merge-gateway$0.60$3.00$0.10—262K262K—
novita$0.60$3.00$0.10—262K262K—
novita-ai · moonshotai$0.60$3.00$0.10—262K262K—
opencode$0.60$3.00$0.10—262K66K—
orcarouter · kimi$0.60$3.00$0.10free262K33K—
poe$0.60$3.00$0.10—128K262K—
vercel · moonshotai$0.60$3.00——256K256K—
wandb · wandb$0.60$3.00$0.10—262K262K—
302ai$0.627$3.30——262K262K—
daoxe$4.00$16.00$0.80—262K262K—

Price history

append-only observations · a listing writes a row only when its price moves

DateHostField$/M
2026-09-02daoxeCache read$0.10
2026-10-22daoxeCache read$0.80
2026-09-02llmgateway-providersCache read$0.07
2026-09-07llmgateway-providersCache read$0.10
2026-09-02nano-gpt · moonshotaiCache read$0.15
2026-10-13nano-gpt · moonshotaiCache read$0.07
2026-09-02nano-gpt · thinkingCache read$0.15
2026-10-13nano-gpt · thinkingCache read$0.07
2026-09-02opencodeCache read$0.08
2026-10-22opencodeCache read$0.10
2026-09-02openrouterCache read$0.10
2026-09-06openrouterCache read$0.07
2026-09-02openrouter · moonshotaiCache read$0.07
2026-10-11openrouter · moonshotaiCache read$0.15
2026-09-02hyperCache write$0.2752
2026-09-02hyperCache write$0.2642
2026-09-05hyperCache write$0.2572
2026-09-08hyperCache write$0.2792
2026-09-02aurikoInput$0.50
2026-10-22aurikoInput$0.45
2026-09-02daoxeInput$0.60
2026-10-22daoxeInput$4.00
2026-09-02hyperInput$0.5504
2026-09-02hyperInput$0.5284
2026-09-05hyperInput$0.5144
2026-09-08hyperInput$0.5584
2026-09-15hyperInput$0.5344
2026-09-16hyperInput$0.5584
2026-09-02llmgatewayInput$0.405
2026-10-22llmgatewayInput$0.574
2026-09-02llmgateway-providersInput$0.45
2026-09-04llmgateway-providersInput$0.574
2026-09-07llmgateway-providersInput$0.574
2026-10-22llmgateway-providersInput$0.60
2026-10-23llmgateway-providersInput$0.574
2026-10-24llmgateway-providersInput$0.60
2026-10-26llmgateway-providersInput$0.574
2026-10-27llmgateway-providersInput$0.60
2026-10-28llmgateway-providersInput$0.574
2026-10-29llmgateway-providersInput$0.60
2026-10-31llmgateway-providersInput$0.574
2026-09-02nano-gpt · moonshotaiInput$0.30
2026-10-13nano-gpt · moonshotaiInput$0.375
2026-09-02nano-gpt · thinkingInput$0.30
2026-10-13nano-gpt · thinkingInput$0.375
2026-09-02openrouterInput$0.60
2026-09-06openrouterInput$0.45
2026-09-02openrouter · moonshotaiInput$0.45
2026-10-11openrouter · moonshotaiInput$0.50
2026-09-02alibaba-cnOutput$2.411
2026-10-21alibaba-cnOutput$3.011
2026-09-02aurikoOutput$2.80
2026-10-22aurikoOutput$2.25
2026-09-02daoxeOutput$3.00
2026-10-22daoxeOutput$16.00
2026-09-02hyperOutput$2.885
2026-09-02hyperOutput$2.785
2026-09-05hyperOutput$2.755
2026-09-08hyperOutput$2.935
2026-09-15hyperOutput$2.815
2026-09-16hyperOutput$2.935
2026-09-02llmgatewayOutput$1.98
2026-10-22llmgatewayOutput$3.011
2026-09-02llmgateway-providersOutput$2.25
2026-09-04llmgateway-providersOutput$3.011
2026-09-07llmgateway-providersOutput$3.011
2026-10-22llmgateway-providersOutput$3.00
2026-10-23llmgateway-providersOutput$3.011
2026-10-24llmgateway-providersOutput$3.00
2026-10-26llmgateway-providersOutput$3.011
2026-10-27llmgateway-providersOutput$3.00
2026-10-28llmgateway-providersOutput$3.011
2026-10-29llmgateway-providersOutput$3.00
2026-10-31llmgateway-providersOutput$3.011
2026-09-02meganova · moonshotaiOutput$2.80
2026-10-22meganova · moonshotaiOutput$2.25
2026-09-02nano-gpt · moonshotaiOutput$1.90
2026-10-13nano-gpt · moonshotaiOutput$2.00
2026-09-02nano-gpt · thinkingOutput$1.90
2026-10-13nano-gpt · thinkingOutput$2.00
2026-09-02openrouterOutput$3.00
2026-09-06openrouterOutput$2.25
2026-09-02openrouter · moonshotaiOutput$2.25
2026-10-11openrouter · moonshotaiOutput$2.50
2026-09-09302aiInput$0.66
2026-10-22302aiInput$0.627
2026-09-11hyperCache read$0.2792
2026-09-15hyperCache read$0.2672
2026-09-16hyperCache read$0.2792

More from Kimi

most-hosted first