Kimi K2.6 (Non-reasoning)
KimiKimiOpen WeightModified MIT License
Описание
Kimi K2.6 is Moonshot AI's open-source, native multimodal agentic model focused on state-of-the-art coding, long-horizon execution, and agent swarm capabilities. It scales horizontally to 300 sub-agents executing 4,000 coordinated steps, dynamically decomposing tasks into parallel, domain-specialized subtasks. K2.6 unifies text, image, and video input with thinking and non-thinking modes, supports a 256K context, and powers proactive 24/7 background agents that manage schedules, execute code, and orchestrate cross-platform operations without human oversight.
Дата выхода
2026-04-20
Параметры
1.0T
Длина контекста
262K
Модальности
image, text, video
Радар способностей
38
general
39
coding
79
reasoning
53
scienceоцен.
60
agents
80
multimodal
Science использует прокси на основе рассуждений, когда специализированные научные бенчмарки недоступны.
Рейтинги
| Домен | #Место | Оценка | Источник |
|---|---|---|---|
| Agents & Tools | 20 | 68.0 | LS |
| Code Ranking | 69 | 67.0 | AA |
| General Ranking | 88 | 70.0 | AA |
| Multimodal Ranking | 22 | 84.0 | LS |
| Reasoning | 62 | 61.0 | LS |
| Science | 90 | 63.0 | AA |
Оценки бенчмарков (LLM Stats)
Agents
BrowseComp
86.3%Сам.
DeepSearchQA
83.0%Сам.
Claw-Eval
80.9%Сам.
WideSearch
80.8%Сам.
OSWorld-Verified
73.1%Сам.
Terminal-Bench 2.0
66.7%Сам.
SWE-Bench Pro
58.6%Сам.
MCP-Mark
55.9%Сам.
Toolathlon
50.0%Сам.
APEX-Agents
27.9%Сам.
Biology
GPQA
90.5%Сам.
SciCode
52.2%Сам.
Code
SWE-Bench Verified
80.2%Сам.
SWE-bench Multilingual
76.7%Сам.
General
LiveCodeBench v6
89.6%Сам.
MMMU-Pro
80.1%Сам.
Math
AIME 2026
96.4%Сам.
MathVision
93.2%Сам.
HMMT Feb 26
92.7%Сам.
IMO-AnswerBench
86.0%Сам.
Humanity's Last Exam
36.4%Сам.
Multimodal
V*
96.9%Сам.
CharXiv-R
86.7%Сам.
BabyVision
68.5%Сам.
Reasoning
OJBench
60.6%Сам.
Индексы оценки AA
Intelligence Index43.0
Coding Index38.4
Tau20.9
Gpqa0.8
Lcr0.6
Ifbench0.4
Scicode0.4
Terminalbench Hard0.4
Hle0.2
Оценки категорий LLM Stats
Vision80
Coding80
Frontend Development80
General80
Math80
Multimodal80
Reasoning80
Search80
Agents70
Biology70
Chemistry70
Code70
Physics70
Tool Calling60
Цены
Цена ввода$0.95 / 1M tokens
Цена вывода$4 / 1M tokens
Смешанная цена (3:1)$1.712 / 1M tokens
Скорость
Токенов/сек28.5 tokens/s
Задержка первого токена1.71s
Время до первого ответа1.71s
Доступные провайдеры
(Внутренние единицы LS)| Провайдер | Цена ввода | Цена вывода |
|---|---|---|
| Novita | 950K | 4.0M |
| Moonshot AI | 950K | 4.0M |