Kimi K3 (Max)
KimiKimiОткрытые весаKimi K3 License · Коммерческое использование
Описание
Kimi K3 is Moonshot AI's flagship open Mixture-of-Experts model for long-horizon coding, knowledge work, and reasoning. It has 2.8 trillion total parameters, activates 16 of 896 experts, and combines Kimi Delta Attention, Attention Residuals, and Stable LatentMoE. The model supports a 1M-token context window, native image and video understanding, tool calling, structured output, and always-on reasoning with max thinking effort at launch.
Дата выхода
2026-07-16
Параметры
2.8T
Длина контекста
1.0M
Модальности
image, text, video
Радар способностей
44
general
74
coding
93
reasoning
72
science
70
agents
80
multimodal
Рейтинги
| Домен | #Место | Оценка | Источник |
|---|---|---|---|
| Агентные возможности | 42 | 52.0 | LS |
| Рейтинг кодинга | 8 | 95.0 | AA |
| Общий рейтинг | 32 | 74.0 | AA |
| Математическое мышление | 47 | 84.0 | LB |
| Мультимодальный рейтинг | 13 | 67.0 | LS |
| Рассуждения | 11 | 91.0 | LB |
| Наука | 12 | 87.0 | AA |
Оценки бенчмарков (LLM Stats)
(LLM Stats (zeroeval))Agents
AA-Briefcase
1548.00 / 3000
Program Bench
77.8%Сам.
DECK-Bench
73.5%Сам.
Toolathlon
73.2%Сам.
Kimi Code Bench v2
72.9%Сам.
OfficeQA Pro
63.3%Сам.
Job Bench
52.9%Сам.
MLS-Bench Lite
48.3%Сам.
PostTrainBench
36.6%Сам.
SpreadsheetBench 2
34.8%Сам.
AutomationBench
30.8%Сам.
Code
FrontierSWE
81.2%Сам.
DeepSWE 1.1
69.0%
DeepSWE
67.5%Сам.
SWE-Marathon
42.0%Сам.
Math
MathVision
97.8%Сам.
Reasoning
GPQANYU + Cohere + Anthropic (2023)
93.5%Сам.
CharXiv-R
91.3%Сам.
BrowseCompOpenAI (2025)
91.2%Сам.
Terminal-Bench 2.1
88.3%Сам.
MCP Atlas
84.2%Сам.
Humanity's Last Exam
56.0%Сам.
APEX-Agents
37.6%Сам.
Search
DeepSearchQA
95.0%Сам.
Vision
OmniDocBench
91.1%Сам.
BabyVision
85.7%Сам.
MMMU-Pro (with tools)
83.4%Сам.
MMMU-Pro
81.6%Сам.
PerceptionBench
58.5%Сам.
WorldVQA
51.0%Сам.
ZEROBench
0.41 / 100Сам.
Индексы оценки AA
(Artificial Analysis)Gpqa(NYU + Cohere + Anthropic (2023))93.5
Lcr(Artificial Analysis)88.7
Terminalbench V2 185.0
Coding Index(Artificial Analysis)76.2
Scicode(UIUC + Argonne National Lab (2024))59.5
Hle(Center for AI Safety + Scale AI (2025))46.9
Tau Banking46.0
Intelligence Index(Artificial Analysis)43.6
Terminalbench V4 012.6
Оценки категорий LLM Stats
(LLM Stats (zeroeval))Productivity100
Agents92
Reasoning78
General60
Physics90
Search90
Biology90
Chemistry90
Math80
Multimodal80
Code70
Tool Calling70
Vision70
Systems40
Цены
Цена ввода$3 / 1M токенов
Цена вывода$15 / 1M токенов
Смешанная цена (3:1)$6 / 1M токенов
Цена чтения кэша$0.169 / 1M токенов
Скорость
Токенов/сек0.0
Задержка первого токена0.00s
Время до первого ответа0.00s
Рейтинг цен провайдеров
Рейтинг цен провайдеров
34 провайдеров
Самый дешевый: DeepInfraСамый дорогой: Tinfoil
ПровайдерВводВывод
1DeepInfraСамый дешевый
$0
$0.00001
2Together
$0
$0.00002
3Moonshot AI
$0
$0.00002
4Fireworks
$0
$0.00002
5Novita
$0
$0.00002
6Kenari
$0.85714
$8.57143
7engy
$1.95
$9.75
8Boundless
$2.3
$11.4
9Vancine
$2.4
$12
10Alibaba (China)
$2.827
$14.133
11Moonshot AI (China)
$2.9879
$14.9395
12KimiОсновной
$3
$15
13302.AI
$3
$15
14Alibaba
$3
$15
15Moark
$3
$15
16OpenCode Go
$3
$15
17Vivgrid
$3
$15
18GitHub Copilot
$3
$15
19OpenCode
$3
$15
20Ollama Cloud
$3
$15
21AIHubMix
$3
$15
22Requesty
$3
$15
23DigitalOcean
$3
$15
24Cortecs
$3
$14.999
25DevPass (LLM Gateway)
$3
$15
26Neuralwatt
$3
$15
27Neon
$3
$15
28EmpirioLabs AI
$3
$15
29Opper
$3
$15
30Melious
$3.19628
$15.98138
31Charm Hyper
$3.2664
$16.332
32Venice AI
$3.75
$18.75
33GreenPT
$3.762
$18.81
34Tinfoil
$4
$20
Сравнение цен разных API-провайдеров для этой модели.