跳转到主要内容

Kimi K2.6 (Non-reasoning)

KimiKimi开源权重Modified MIT License

描述

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.

发布日期
2026-04-20
参数规模
1.0T
上下文长度
262K
支持模态
image, text, video

能力雷达图

32
general
40
coding
79
reasoning
53
science
60
agents
80
multimodal

排行榜排名

领域#排名分数来源
智能体能力模型榜28
56.0
LS
代码能力榜130
68.0
AA
通用能力榜142
63.0
AA
数学推理27
84.0
LB
多模态榜14
63.0
LS
推理能力26
79.0
LB
科学能力168
60.0
AA

基准测试分数 (LLM Stats)

(LLM Stats (zeroeval))

Agents

BrowseCompOpenAI (2025)86.3%自报
DeepSearchQA83.0%自报
Claw-Eval80.9%自报
WideSearch80.8%自报
OSWorld-Verified73.1%自报
Terminal-Bench 2.0Stanford × Laude Institute (2026)66.7%自报
SWE-Bench ProPrinceton NLP (2024)58.6%自报
MCP-Mark55.9%自报
Toolathlon50.0%自报
Finance Agent v244.9%
APEX-Agents27.9%自报
FrontierSWE27.0%

Biology

GPQANYU + Cohere + Anthropic (2023)90.5%自报
SciCode52.2%自报

Code

SWE-Bench Verified80.2%自报
SWE-bench Multilingual76.7%自报

General

LiveCodeBench v689.6%自报
MMMU-Pro80.1%自报
LiveBench72.2%

Math

AIME 202696.4%自报
MathVision93.2%自报
HMMT Feb 2692.7%自报
IMO-AnswerBench86.0%自报
Humanity's Last Exam36.4%自报

Multimodal

V*96.9%自报
CharXiv-R86.7%自报
BabyVision68.5%自报

Reasoning

OJBench60.6%自报

AA 评测指数

(Artificial Analysis)
Intelligence Index(Artificial Analysis)
35.4
Tau2(Sierra + U Toronto + Vector Institute (2025))
0.9
Gpqa(NYU + Cohere + Anthropic (2023))
0.8
Lcr(Artificial Analysis)
0.7
Ifbench(Google Research (2023))
0.4
Scicode(UIUC + Argonne National Lab (2024))
0.4
Terminalbench Hard(Stanford × Laude Institute (2026))
0.4
Hle(Center for AI Safety + Scale AI (2025))
0.2

LLM Stats 分类评分

(LLM Stats (zeroeval))
Math
80
Multimodal
80
Search
80
Frontend Development
80
Vision
80
Physics
70
Reasoning
70
General
70
Biology
70
Chemistry
70
Agents
60
Code
60
Tool Calling
60
Finance
40

定价

输入价格$0.95 / 1M tokens
输出价格$4 / 1M tokens
混合价格(3:1)$1.712 / 1M tokens
缓存读取价格$0.0944 / 1M tokens

速度

Tokens/秒0.0
首Token延迟0.00s
首回答延迟0.00s

供应商价格排行

供应商价格排行

17 个供应商

最便宜: DeepInfra最贵: TensorX
供应商输入输出
1DeepInfra最便宜
$0
$0
2Fireworks
$0
$0
3Novita
$0
$0
4Moonshot AI
$0
$0
5Together
$0
$0
6NanoGPT
$0.5
$2.6
7OpenRouter
$0.5605
$2.36
8Lilac
$0.7
$3.5
9FastRouter
$0.75
$3.5
10NovitaAI
$0.8
$3.4
11Kilo Gateway
$0.8
$3.4
12Kimi主要
$0.95
$4
13ZenMux
$0.95
$4
14Vercel AI Gateway
$0.95
$4
15Ambient
$0.95
$4
16Ofox
$0.95
$4
17TensorX
$1
$4

比较该模型在不同 API 供应商之间的定价。

外部链接