소형 오픈 웨이트 AI 모델 비교(4B~40B)

파라미터 수가 4B~40B개인 오픈 웨이트 AI 모델입니다.

웨이트를 다운로드할 수 있는 모델을 오픈 웨이트 모델(흔히 오픈 소스라고도 함)로 간주합니다. 자체 인프라에서 호스팅할 수 있으며 미세 조정 등을 통해 모델을 맞춤 설정할 수 있습니다.

방법론을 비롯한 자세한 내용은 FAQ에서 확인하세요.

Alibaba 로고Qwen3.8 27B (xhigh)Alibaba 로고Qwen3.8 27B (medium)은 파라미터 수가 4B~40B개인 소형 오픈 웨이트 모델 중 지능이 가장 높으며 Alibaba 로고Qwen3.8 27B (low)MBZUAI Institute of Foundation Models 로고K2 Horizon MoVA 36B A4B이 뒤를 잇습니다.

주요 내용

Artificial Analysis Openness Index · Higher is better
Updated
Artificial Analysis Intelligence Index · Higher is better
학습 가능한 파라미터 수(단위: 십억 개)

개방성

Artificial Analysis Openness Index: Score

Openness Index assesses model openness on a 0 to 100 normalized scale (higher is more open)

지능

Artificial Analysis Intelligence Index

Artificial Analysis Intelligence Index v4.3 incorporates 10 evaluations: AA-Briefcase, GDPval-AA v2, AutomationBench-AA, Terminal-Bench v4.0, SciCode, Humanity's Last Exam, GDP.pdf, CritPt, AA-Omniscience, AA-LCR v1.1
Estimate (independent evaluation forthcoming)

Intelligence Evaluations

Intelligence evaluations measured independently by Artificial Analysis · Higher is better
See more

Agentic knowledge work, (Elo-500)/2000

Agentic real-world work tasks, (Elo-500)/2000

Agentic SaaS workflows

Agentic coding & terminal use

Coding

Reasoning & knowledge

Professional document reasoning, All-pass

Physics reasoning

Long context reasoning

Legal agentic work, criterion pass rate

Agentic business operations

Quantitative analysis on spreadsheets & documents

No data available

Instruction following

Agentic tool use

Long-horizon agentic tasks

No data available

Kubernetes incident root-cause analysis

Visual reasoning

크기

Model Size: Total and Active Parameters

Comparison between total model parameters and parameters active during inference

Intelligence Index vs. Active Parameters

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

Intelligence Index vs. Total Parameters

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

컨텍스트 창

Context Window

Context window: tokens limit · Higher is better

상세 정보

웨이트
제공업체 벤치마크
Qwen3.8 27B (xhigh)
Alibaba 로고Alibaba
34
27B
256k
$0.4
45
DeepInfraSelf-hostedCoreWeave
+5
Qwen3.8 27B (medium)
Alibaba 로고Alibaba
28
27B
256k
$0.4
51
Alibaba CloudMultiverse Computing
Qwen3.8 27B (low)
Alibaba 로고Alibaba
26
27B
256k
$0.4
50
Alibaba CloudMultiverse Computing
K2 Horizon MoVA 36B A4B
MBZUAI Institute of Foundation Models 로고MBZUAI Institute of Foundation Models
26
36B
추론 시 4B 활성
524k
-
-
-
Qwen3.8 27B (Non-reasoning)
Alibaba 로고Alibaba
22
27B
256k
$0.4
50
Alibaba CloudMultiverse Computing
Qwen3.6 27B (Reasoning)
Alibaba 로고Alibaba
22
27.8B
262k
$0.9
57
DeepInfraGroqNovita
+2
G9v3-39A5B
AI9Stars 로고AI9Stars
22
39B
추론 시 5B 활성
131k
-
-
AI9Stars
K2 Horizon 7B
MBZUAI Institute of Foundation Models 로고MBZUAI Institute of Foundation Models
21
7B
524k
-
-
-
Qwen3.6 27B (Non-reasoning)
Alibaba 로고Alibaba
20
27.8B
262k
$0.9
58
GroqCoreWeaveDeepInfra
+2
Qwen3.6 35B A3B (Reasoning)
Alibaba 로고Alibaba
19
36B
추론 시 3B 활성
262k
$0.6
123
Alibaba CloudSiliconFlowScaleway
+5
Muse Glimmer (high)
Meta 로고Meta
18
30B
131k
$0.2
90
Together AIDeepInfraFireworks
Gemma 4 26B A4B (Reasoning)
Google 로고Google
17
25.2B
추론 시 3.8B 활성
256k
$0.1
-
MakoraNovitaGoogle
+5