Qwen3.5 35B A3B (Non-reasoning)

AlibabaQwenOpen WeightApache 2.0 · Usage Commercial

Description

Qwen3.5-35B-A3B is a multimodal Mixture-of-Experts model with 35 billion total parameters and 3 billion activated parameters. It combines strong reasoning, coding, agentic, and visual understanding performance with production-friendly efficiency and a native 262K context window.

Date de sortie

2026-02-24

Paramètres

35.0B

Longueur du contexte

262K

Modalités

audio, image, text, video

Radar de capacités

general

coding

reasoning

science

agents

multimodal

Classements

Domaine	#Rang	Score	Source
Capacité agentique	96	40.0	LS
Classement codage	256	46.0	AA
Classement général	226	52.0	AA
Classement multimodal	45	47.0	LS
Science	227	52.0	AA

Scores de benchmarks (LLM Stats)

(LLM Stats (zeroeval))

3d

SUNRGBD

0.33 / 100Aut.

Hypersim

0.13 / 100Aut.

Agents

t2-bench

81.2%Aut.

AndroidWorld_SR

71.1%Aut.

BFCL-V4

67.3%Aut.

BrowseCompOpenAI (2025)

61.0%Aut.

FullStackBench en

58.1%Aut.

WideSearch

57.1%Aut.

TIR-Bench

55.5%Aut.

FullStackBench zh

55.0%Aut.

OSWorld-Verified

54.5%Aut.

Terminal-Bench 2.0Stanford × Laude Institute (2026)

40.5%Aut.

VITA-Bench

31.9%Aut.

DeepPlanning

22.8%Aut.

Biology

GPQANYU + Cohere + Anthropic (2023)

84.2%Aut.

Chemistry

SuperGPQA

63.4%Aut.

Code

SWE-Bench Verified

69.2%Aut.

Communication

Multi-Challenge

60.0%Aut.

Embodied

EmbSpatialBench

0.83 / 100Aut.

Finance

MMLU-Pro

85.3%Aut.

MMLU-ProX

81.0%Aut.

General

MMLU-Redux

93.3%Aut.

IFEvalGoogle Research (2023)

91.9%Aut.

C-Eval

90.2%Aut.

MAXIFE

86.6%Aut.

Global PIQA

86.6%Aut.

MMMLU

85.2%Aut.

MMStar

81.9%Aut.

MMMU

81.4%Aut.

Include

79.7%Aut.

MMMU-Pro

75.1%Aut.

LiveCodeBench v6

74.6%Aut.

IFBench

70.2%Aut.

LongBench v2

59.0%Aut.

SimpleVQA

0.58 / 100Aut.

NOVA-63

57.1%Aut.

Grounding

RefCOCO-avg

0.89 / 100Aut.

ScreenSpot Pro

68.6%Aut.

RefSpatialBench

0.64 / 100Aut.

Healthcare

VideoMMMU

80.4%Aut.

SlakeVQA

78.7%Aut.

PMC-VQA

62.0%Aut.

MedXpertQA

61.4%Aut.

Image To Text

OCRBench

91.0%Aut.

Language

LingoQA

79.2%Aut.

WMT24++

76.3%Aut.

Long Context

MLVU

85.6%Aut.

LVBench

71.4%Aut.

MMLongBench-Doc

0.59 / 100Aut.

AA-LCR

58.5%Aut.

Math

HMMT25

89.2%Aut.

HMMT 2025

89.0%Aut.

MathVista-Mini

86.2%Aut.

DynaMath

85.0%Aut.

MathVision

83.9%Aut.

CodeForces

0.82 / 3000Aut.

PolyMATH

64.4%Aut.

Humanity's Last Exam

47.4%Aut.

Multimodal

VLMsAreBlind

97.0%Aut.

92.7%Aut.

AI2D

92.6%Aut.

MMBench-V1.1

91.5%Aut.

OmniDocBench 1.5

89.3%Aut.

VideoMME w sub.

86.6%Aut.

VideoMME w/o sub.

82.5%Aut.

CC-OCR

80.7%Aut.

CharXiv-R

77.5%Aut.

MVBench

74.8%Aut.

MMVU

72.3%Aut.

BabyVision

38.4%Aut.

ZEROBench-Sub

0.34 / 100Aut.

Nuscene

14.6%Aut.

ZEROBench

0.08 / 100Aut.

Reasoning

CountBench

0.98 / 100Aut.

BrowseComp-zh

69.5%Aut.

Hallusion Bench

67.9%Aut.

ERQA

64.8%Aut.

Seal-0

41.4%Aut.

OJBench

36.0%Aut.

Spatial Reasoning

RealWorldQA

84.1%Aut.

Vision

ODinW

42.6%Aut.

Indices d'évaluation AA

(Artificial Analysis)

Coding Index(Artificial Analysis)

37.0

Intelligence Index(Artificial Analysis)

24.3

Tau2(Sierra + U Toronto + Vector Institute (2025))

0.9

Gpqa(NYU + Cohere + Anthropic (2023))

0.8

Lcr(Artificial Analysis)

0.6

Ifbench(Google Research (2023))

0.4

Terminalbench V2 1

0.4

Scicode(UIUC + Argonne National Lab (2024))

0.3

Hle(Center for AI Safety + Scale AI (2025))

0.1

Terminalbench Hard(Stanford × Laude Institute (2026))

0.1

Tau Banking

0.0

Scores par catégorie LLM Stats

(LLM Stats (zeroeval))

Legal

Math

Physics

Structured Output

Image To Text

Instruction Following

Language

Embodied

Finance

Biology

Text-to-image

Video

Long Context

Multimodal

Reasoning

Spatial Reasoning

Frontend Development

General

Grounding

Healthcare

Chemistry

Vision

Code

Communication

Economics

Tool Calling

Agents

Spatial

Tarification

Prix d'entrée$0.25 / 1M tokens

Prix de sortie$2 / 1M tokens

Prix mixte (3:1)$0.688 / 1M tokens

Vitesse

Tokens/sec161.3

Délai du premier token1.21s

Temps de réponse1.21s

Classement des Prix par Fournisseur

1 fournisseurs

FournisseurEntréeSortie

1AlibabaPRINCIPAL

$0.25

Comparer les prix entre différents fournisseurs API pour ce modèle.

Sources externes

LLM Stats Artificial Analysis