मुख्य सामग्री पर जाएं
AITier

Qwen3.5 35B A3B (Non-reasoning)

AlibabaQwenOpen WeightApache 2.0 · Commercial OK

विवरण

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.

रिलीज़ तिथि
2026-02-24
पैरामीटर
35.0B
संदर्भ लंबाई
262K
मोडैलिटीज़
image, text, video

क्षमता रडार

27
general
19
coding
82
reasoning
50
scienceअनुमानित
60
agents
80
multimodal

समर्पित विज्ञान बेंचमार्क उपलब्ध न होने पर Science तर्क प्रॉक्सी का उपयोग करके अनुमान लगाता है।

रैंकिंग

डोमेन#रैंकस्कोरस्रोत
Agents & Tools53
56.0
LS
Code Ranking227
37.0
AA
General Ranking157
57.0
AA
Multimodal Ranking62
69.0
LS
Reasoning52
68.0
LS
Science144
54.0
AA

बेंचमार्क स्कोर (LLM Stats)

3d

SUNRGBD0.33 / 100स्वयं
Hypersim0.13 / 100स्वयं

Agents

t2-bench81.2%स्वयं
AndroidWorld_SR71.1%स्वयं
BFCL-V467.3%स्वयं
BrowseComp61.0%स्वयं
FullStackBench en58.1%स्वयं
WideSearch57.1%स्वयं
TIR-Bench55.5%स्वयं
FullStackBench zh55.0%स्वयं
OSWorld-Verified54.5%स्वयं
Terminal-Bench 2.040.5%स्वयं
VITA-Bench31.9%स्वयं
DeepPlanning22.8%स्वयं

Biology

GPQA84.2%स्वयं

Chemistry

SuperGPQA63.4%स्वयं

Code

SWE-Bench Verified69.2%स्वयं

Communication

Multi-Challenge60.0%स्वयं

Embodied

EmbSpatialBench0.83 / 100स्वयं

Finance

MMLU-Pro85.3%स्वयं
MMLU-ProX81.0%स्वयं

General

MMLU-Redux93.3%स्वयं
IFEval91.9%स्वयं
C-Eval90.2%स्वयं
MAXIFE86.6%स्वयं
Global PIQA86.6%स्वयं
MMMLU85.2%स्वयं
MMStar81.9%स्वयं
MMMU81.4%स्वयं
Include79.7%स्वयं
MMMU-Pro75.1%स्वयं
LiveCodeBench v674.6%स्वयं
IFBench70.2%स्वयं
LongBench v259.0%स्वयं
SimpleVQA0.58 / 100स्वयं
NOVA-6357.1%स्वयं

Grounding

RefCOCO-avg0.89 / 100स्वयं
ScreenSpot Pro68.6%स्वयं
RefSpatialBench0.64 / 100स्वयं

Healthcare

VideoMMMU80.4%स्वयं
SlakeVQA78.7%स्वयं
PMC-VQA62.0%स्वयं
MedXpertQA61.4%स्वयं

Image To Text

OCRBench91.0%स्वयं

Language

LingoQA79.2%स्वयं
WMT24++76.3%स्वयं

Long Context

MLVU85.6%स्वयं
LVBench71.4%स्वयं
MMLongBench-Doc0.59 / 100स्वयं
AA-LCR58.5%स्वयं

Math

HMMT2589.2%स्वयं
HMMT 202589.0%स्वयं
MathVista-Mini86.2%स्वयं
DynaMath85.0%स्वयं
MathVision83.9%स्वयं
CodeForces0.82 / 3000स्वयं
PolyMATH64.4%स्वयं
Humanity's Last Exam47.4%स्वयं

Multimodal

VLMsAreBlind97.0%स्वयं
V*92.7%स्वयं
AI2D92.6%स्वयं
MMBench-V1.191.5%स्वयं
OmniDocBench 1.589.3%स्वयं
VideoMME w sub.86.6%स्वयं
VideoMME w/o sub.82.5%स्वयं
CC-OCR80.7%स्वयं
CharXiv-R77.5%स्वयं
MVBench74.8%स्वयं
MMVU72.3%स्वयं
BabyVision38.4%स्वयं
ZEROBench-Sub0.34 / 100स्वयं
Nuscene14.6%स्वयं
ZEROBench0.08 / 100स्वयं

Reasoning

CountBench0.98 / 100स्वयं
BrowseComp-zh69.5%स्वयं
Hallusion Bench67.9%स्वयं
ERQA64.8%स्वयं
Seal-041.4%स्वयं
OJBench36.0%स्वयं

Spatial Reasoning

RealWorldQA84.1%स्वयं

Vision

ODinW42.6%स्वयं

AA मूल्यांकन सूचकांक

Intelligence Index
30.7
Coding Index
16.8
Tau2
0.9
Gpqa
0.8
Lcr
0.6
Ifbench
0.4
Scicode
0.3
Hle
0.1
Terminalbench Hard
0.1

LLM Stats श्रेणी स्कोर

Structured Output
80
Text-to-image
80
Video
80
Biology
80
Embodied
80
Finance
80
General
80
Image To Text
80
Instruction Following
80
Language
80
Legal
80
Math
80
Physics
80
Spatial Reasoning
70
Vision
70
Chemistry
70
Frontend Development
70
Grounding
70
Healthcare
70
Long Context
70
Multimodal
70
Reasoning
70
Tool Calling
60
Code
60
Communication
60
Economics
60
Search
60
Agents
50
3d
20
Spatial
10

मूल्य निर्धारण

इनपुट मूल्य$0.25 / 1M tokens
आउटपुट मूल्य$2 / 1M tokens
मिश्रित मूल्य (3:1)$0.688 / 1M tokens

गति

टोकन/सेकंड138.7 tokens/s
पहले टोकन में देरी1.15s
पहले उत्तर में देरी1.15s

उपलब्ध प्रदाता

(LS आंतरिक इकाइयाँ)
प्रदाताइनपुट मूल्यआउटपुट मूल्य
Novita250K2.0M

बाहरी लिंक

LLM Stats