Kimi K2.5

Release Date: 1/26/2026

Benchmarked by Vals Logo
Vals Index

Accuracy (Vals Index)

59.61% ± 1.99

Latency (Vals Index)

341.36s

Cost/Test (Vals Index)

$0.12

Context Window

262k

Max Output Tokens

128k

Input Modality

Hyperparameter settings
Default Provider : Moonshot AI

Temperature

1

Top P

0.95

Top K

Default

Max Output Tokens

128,000

Benchmarks

Accuracy

Rankings

Vals Index

0.0%

± 1.99

12/ 34

Vals Multimodal Index

0.0%

± 1.56

9/ 25

CaseLaw (v2)

0.0%

± 0.70

22/ 41

CorpFin

0.0%

± 0.92

1/ 92

Finance Agent (v1.1)

0.0%

± 2.85

18/ 40

MedCode

0.0%

± 2.12

24/ 47

MedScribe

0.0%

± 1.99

27/ 47

MortgageTax

0.0%

± 0.93

16/ 66

SAGE

0.0%

± 3.36

5/ 46

TaxEval (v2)

0.0%

± 0.85

15/ 100

Vibe Code Bench

0.0%

± 3.26

14/ 22

AIME

0.0%

± 0.34

9/ 92

GPQA

0.0%

± 2.13

19/ 95

IOI

0.0%

± 6.08

15/ 50

LiveCodeBench

0.0%

± 1.03

18/ 101

MedQA

0.0%

± 0.21

17/ 95

MMMU

0.0%

± 0.87

6/ 63

SWE-bench

0.0%

± 2.05

18/ 37

Terminal-Bench 2.0

0.0%

± 5.23

18/ 46

Academic Benchmarks
Proprietary Benchmarks (contact us to get access)

Join our mailing list to receive benchmark updates

Model benchmarks are seriously lacking. With Vals AI, we report how language models perform on the industry-specific tasks where they will be used.

By subscribing, I agree to Vals' Privacy Policy.