开放权重模型比较

从质量、性能、推理速度、上下文窗口、参数量和许可证详情等关键指标比较和分析开放权重 AI 模型。

如果模型权重可供下载,我们便将其视为开放权重模型。这样用户就可以在自己的基础设施上自行托管,并通过微调等方式定制模型。

如需了解我们方法论的更多详情,请参阅常见问题。

Kimi 标志Kimi K3.1 Pro 和 DeepSeek 标志DeepSeek V4.1 Flash 1017 (max) 是智能得分最高的开放权重模型,其次是 Z AI 标志GLM-5.3 (max) 和 Thinking Machines 标志Inkling 2 Small。
Artificial Analysis Openness Index · Higher is better
Artificial Analysis Intelligence Index · Higher is better
可训练参数(十亿)

开放性

Artificial Analysis Openness Index:得分

Openness Index 以 0 至 100 的标准化评分衡量模型的开放程度(分数越高,越开放)

开放权重模型进展

开放权重与专有模型:Artificial Analysis Progress Index 变化趋势

Artificial Analysis Progress Index v1 is an item response theory estimate of general capability across Artificial Analysis evaluations · Highest-scoring effort variant shown
Nov 2022Nov 2026

开放权重:各实验室 Artificial Analysis Progress Index 随时间的变化

Artificial Analysis Progress Index v1 is an item response theory estimate of general capability across Artificial Analysis evaluations · Highest-scoring effort variant shown

开放权重:各规模 Artificial Analysis Progress Index 随时间的变化

Artificial Analysis Progress Index v1 is an item response theory estimate of general capability across Artificial Analysis evaluations · Highest-scoring effort variant shown

智能

Artificial Analysis Intelligence Index

Artificial Analysis Intelligence Index v5.0.1 incorporates 9 evaluations: AA-Briefcase v1.1, AutomationBench-AA, Terminal-Bench 4.0, FrontierCode-AA 2.0, AA-Omniscience v1.1, GDP.pdf, HLE-Diamond, CritPt v2 (Hard Subset), and Terminal-Bench-Science 0.1

Evaluations

Evaluations measured independently by Artificial Analysis · AA-PersonalityBench scores are percentile ranks from 1 to 100
查看更多

Agentic knowledge work, Elo

Agentic SaaS workflows

Agentic scientific research workflows in a terminal

Agentic software engineering in open-source repositories

Reasoning & knowledge

Professional document reasoning, All-pass

Agentic coding & terminal use

Physics reasoning

Quantitative analysis on spreadsheets & documents

Legal agentic work, Hallucination-Gated All-Pass Rate

Agentic business operations

Agentic real-world work tasks, Elo

Agentic interactive 3D experiences, Elo

Product and brand videos designed in code, Elo

Kubernetes incident root-cause analysis

Medical long context reasoning

Visual reasoning

Rich, varied, well-crafted language

Well-structured, scannable, answer-first

Truthful, acknowledges uncertainty, non-sycophantic

Use of filler and generic language

规模

按模型规模划分的 Intelligence Index

Artificial Analysis Intelligence Index v5.0.1 incorporates 9 evaluations: AA-Briefcase v1.1, AutomationBench-AA, Terminal-Bench 4.0, FrontierCode-AA 2.0, AA-Omniscience v1.1, GDP.pdf, HLE-Diamond, CritPt v2 (Hard Subset), and Terminal-Bench-Science 0.1

模型规模:总参数量与活跃参数量

Comparison between total model parameters and parameters active during inference (billions)

Intelligence Index 与活跃参数

Artificial Analysis Intelligence Index · Active parameters at inference time (billions)
Most attractive quadrant
Pareto line

Intelligence Index 与总参数量

Artificial Analysis Intelligence Index · Size in parameters (billions)
Most attractive quadrant
Pareto line

上下文窗口

上下文窗口

Context window: tokens limit · Higher is better

更多详情

权重
服务商基准测试
Kimi K3.1 Pro
Kimi 标志Kimi
46
2.8T
推理时启用 104B 个参数
1M
$1.54
38.7
31
DatabricksIncoKimi
+15
DeepSeek V4.1 Flash 1017 (Max)
DeepSeek 标志DeepSeek
33
552B
推理时启用 16B 个参数
1M
$0.18
13.1
177
BasetenNovitaBaseten
+23
GLM-5.3 (Max)
Z AI 标志Z AI
32
753B
推理时启用 40B 个参数
1M
$0.90
22.2
89
ModalZaiSambaNova
+26
Inkling 2 Small
Thinking Machines 标志Thinking Machines
31
594B
推理时启用 17B 个参数
1M
$0.18
21.4
117
DeepInfraNebiusTogether AIThinking Machines
MiMo-V2.6-Pro
Xiaomi 标志Xiaomi
30
1T
推理时启用 42B 个参数
1M
$0.18
35.8
56
XiaomiDeepInfraPrimaLabsGMI
Ling 3.1 Flash 1019
InclusionAI 标志InclusionAI
30
560B
推理时启用 25B 个参数
1M
$0.19
-
-
InclusionAI
Mistral Large 4 (High)
Mistral 标志Mistral
26
1.1T
推理时启用 52B 个参数
1M
$0.40
18.4
74
Mistral
Qwen3.8 27B (Xhigh)
Alibaba 标志Alibaba
20
27B
256k
$0.47
38.8
47
DeepInfraCoreWeaveAlibaba Cloud
+7
K2 Horizon 375B A23B
Institute of Foundation Models 标志Institute of Foundation Models
18
375B
推理时启用 23B 个参数
524k
$0.00
4.7
224
Institute of Foundation Models
Beam 1.0
Reflection 标志Reflection
18
501B
推理时启用 23B 个参数
1M
$0.23
2.2
257
BasetenCoreWeaveTogether AI
+2
Nemotron 3.5 Super VL
NVIDIA 标志NVIDIA
14
120B
推理时启用 12B 个参数
1M
$0.13
4.5
300
PrimaLabsDeepInfraBitdeer AI
+6