Comparação de modelos médios de IA de pesos abertos (40B-150B)

Name: Abertura
Creator: Artificial Analysis
License: https://artificialanalysis.ai/docs/legal/Terms-of-Use.pdf

Modelos de IA de pesos abertos com 40B a 150B parâmetros.

Os modelos são considerados de pesos abertos (também chamados frequentemente de código aberto) quando seus pesos estão disponíveis para download. Isso permite a auto-hospedagem em sua própria infraestrutura e a personalização do modelo, por exemplo, por meio de ajuste fino.

Para mais detalhes sobre nossa metodologia, consulte as perguntas frequentes.

Nemotron 3.5 Super VL e

Ling-3.0-flash-VL são os modelos com maior inteligência entre os modelos médios de pesos abertos, definidos como aqueles com 40B-150B parâmetros, seguidos por

Ling-3.0-flash-Fin e

Nemotron 3.5 Bolt 74B-A9B.

Abertura

Artificial Analysis Openness Index · Higher is better

Inteligência

Artificial Analysis Intelligence Index · Higher is better

Parâmetros totais

Parâmetros treináveis em bilhões

Abertura

Artificial Analysis Openness Index: pontuação

O Openness Index avalia a abertura dos modelos em uma escala normalizada de 0 a 100 (quanto maior, mais aberto)

Inteligência

Artificial Analysis Intelligence Index

Artificial Analysis Intelligence Index v4.3.2 incorporates 10 evaluations: AA-Briefcase v1.1, GDPval-AA v2.1, AutomationBench-AA, Terminal-Bench 4.0, SciCode, Humanity's Last Exam, GDP.pdf, CritPt, AA-Omniscience, AA-LCR v1.1

Estimate (independent evaluation forthcoming)

Artificial Analysis Intelligence Index v4.3.2 includes: AA-Briefcase v1.1, GDPval-AA v2.1, AutomationBench-AA, Terminal-Bench 4.0, SciCode, Humanity's Last Exam, GDP.pdf, CritPt, AA-Omniscience, AA-LCR v1.1. See Intelligence Index methodology for further details, including a breakdown of each evaluation and how we run them.

Evaluations

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

Ver mais

AA-Briefcase v1.1

Agentic knowledge work, Elo

AA-Omniscience Accuracy

Knowledge

AA-Omniscience Non-Hallucination Rate

1 - hallucination rate

AutomationBench-AA

Agentic SaaS workflows

GDPval-AA v2.1

Agentic real-world work tasks, Elo

Humanity's Last Exam

Reasoning & knowledge

GDP.pdf

Professional document reasoning, All-pass

Terminal-Bench 4.0

Agentic coding & terminal use

AA-LCR v1.1

Long context reasoning

SciCodeUnder review

Coding

CritPt

Physics reasoning

Terminal-Bench-Science 0.1New

Agentic scientific research workflows in a terminal

FrontierCode 1.1New

Agentic software engineering in open-source repositories

AA-AnalystAgent

Quantitative analysis on spreadsheets & documents

Harvey LAB-AA v1.1Updated

Legal agentic work, Hallucination-Gated All-Pass Rate

No data available

EnterpriseOps-Gym-AA

Agentic business operations

AA-3DNew

Agentic interactive 3D experiences, Elo

No data available

Code2Video-AANew

Product and brand videos designed in code, Elo

No data available

ITBench-AA

Kubernetes incident root-cause analysis

MLCR-AAUpdated

Medical long context reasoning

MMMU-Pro

Visual reasoning

AA-PersonalityBench: Expressiveness ScoreNew

Rich, varied, well-crafted language

AA-PersonalityBench: Organization ScoreNew

Well-structured, scannable, answer-first

AA-PersonalityBench: Transparency ScoreNew

Truthful, acknowledges uncertainty, non-sycophantic

AA-PersonalityBench: Slop ScoreNew

Use of filler and generic language

While model intelligence generally translates across use cases, specific evaluations may be more relevant for certain use cases.

Tamanho

Tamanho do modelo: parâmetros totais e ativos

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

The total number of trainable weights and biases in the model, expressed in billions. These parameters are learned during training and determine the model's ability to process and generate responses.

The number of parameters actually executed during each inference forward pass, expressed in billions. For Mixture of Experts (MoE) models, a routing mechanism selects a subset of experts per token, resulting in fewer active than total parameters. Dense models use all parameters, so active equals total.

Intelligence Index vs. parâmetros ativos

Artificial Analysis Intelligence Index · Active parameters at inference time (billions)

Most attractive quadrant

Pareto line

Intelligence Index vs. parâmetros totais

Artificial Analysis Intelligence Index · Size in parameters (billions)

Most attractive quadrant

Pareto line

The total number of trainable weights and biases in the model, expressed in billions. These parameters are learned during training and determine the model's ability to process and generate responses.

Janela de contexto

Context window: tokens limit · Higher is better

Larger context windows are relevant to RAG (Retrieval Augmented Generation) LLM workflows which typically involve reasoning and information retrieval of large amounts of data.

Maximum number of combined input & output tokens. Output tokens commonly have a significantly lower limit (varied by model).

Mais detalhes

							Pesos		Benchmarks de provedores
Nemotron 3.5 Super VL	NVIDIA	26	120B 12B ativos durante a inferência	1M	$0.1	454	🤗	+5	Ver
Ling-3.0-flash-VL	InclusionAI	25	124B 5.5B ativos durante a inferência	262k	-	-	🤗	-	Ver
Ling-3.0-flash-Fin	InclusionAI	23	124B 5.1B ativos durante a inferência	262k	-	-	🤗	-	Ver
Nemotron 3.5 Bolt 74B-A9B	NVIDIA	21	74B 9B ativos durante a inferência	262k	-	-	🤗	-	Ver
Qwen3.5 122B A10B (Non-reasoning)	Alibaba	18	125B 10B ativos durante a inferência	262k	$0.7	145	🤗		Ver
Qwen3.5 122B A10B (Reasoning)	Alibaba	16	125B 10B ativos durante a inferência	262k	$0.7	131	🤗		Ver
Mistral Medium 3.5	Mistral	14	128B	256k	$1.2	170	🤗		Ver
HyperNova 60B 2605 (High, Based on gpt-oss-120b)	Multiverse Computing	12	58.7B 4.8B ativos durante a inferência	131k	-	-	🤗	-	Ver
gpt-oss-120b (High)	OpenAI	12	117B 5.1B ativos durante a inferência	131k	$0.2	241	🤗	+14	Ver
K2 Think V2	Institute of Foundation Models	11	70B	262k	-	-	🤗	-	Ver
LongCat Flash Lite	LongCat	11	68.5B 3B ativos durante a inferência	256k	-	-	🤗	-	Ver
Mistral Small 4 (Reasoning)	Mistral	11	119B 6.5B ativos durante a inferência	256k	$0.1	179	🤗		Ver

Comparação de modelos médios de IA de pesos abertos (40B-150B)

Abertura

Inteligência

Parâmetros totais

Abertura

Artificial Analysis Openness Index: pontuação

Inteligência

Artificial Analysis Intelligence Index

Artificial Analysis Intelligence Index

Evaluations

Intelligence Evaluation Relevance

Artificial Analysis Intelligence Index

Tamanho

Tamanho do modelo: parâmetros totais e ativos

Total Parameters

Active Parameters at Inference Time

Intelligence Index vs. parâmetros ativos

Artificial Analysis Intelligence Index

Active Parameters at Inference Time

Intelligence Index vs. parâmetros totais

Artificial Analysis Intelligence Index

Total Parameters

Janela de contexto

Janela de contexto

Context Window for RAG

Context Window

Mais detalhes