Перейти к основному содержанию
AITier

Qwen3.5 122B A10B (Non-reasoning)

AlibabaQwenOpen WeightApache 2.0 · Commercial OK

Описание

Qwen3.5-122B-A10B is a multimodal Mixture-of-Experts model with 122 billion total parameters and 10 billion activated parameters. It combines strong reasoning, coding, long-context, and visual understanding performance with production-friendly efficiency and a native 262K context window.

Дата выхода
2026-02-24
Параметры
122.0B
Длина контекста
262K
Модальности
image, text, video

Радар способностей

31
general
32
coding
83
reasoning
53
scienceоцен.
60
agents
80
multimodal

Science использует прокси на основе рассуждений, когда специализированные научные бенчмарки недоступны.

Рейтинги

Домен#МестоОценкаИсточник
Agents & Tools45
58.0
LS
Code Ranking116
57.0
AA
General Ranking121
63.0
AA
Multimodal Ranking61
70.0
LS
Reasoning53
68.0
LS
Science108
59.0
AA

Оценки бенчмарков (LLM Stats)

3d

SUNRGBD0.36 / 100Сам.
Hypersim0.13 / 100Сам.

Agents

t2-bench79.5%Сам.
BFCL-V472.2%Сам.
AndroidWorld_SR66.4%Сам.
BrowseComp63.8%Сам.
FullStackBench en62.6%Сам.
WideSearch60.5%Сам.
FullStackBench zh58.7%Сам.
OSWorld-Verified58.0%Сам.
TIR-Bench53.2%Сам.
Terminal-Bench 2.049.4%Сам.
VITA-Bench33.6%Сам.
DeepPlanning24.1%Сам.

Biology

GPQA86.6%Сам.

Chemistry

SuperGPQA67.1%Сам.

Code

SWE-Bench Verified72.0%Сам.

Communication

Multi-Challenge61.5%Сам.

Embodied

EmbSpatialBench0.84 / 100Сам.

Finance

MMLU-Pro86.7%Сам.
MMLU-ProX82.2%Сам.

General

MMLU-Redux94.0%Сам.
IFEval93.4%Сам.
C-Eval91.9%Сам.
Global PIQA88.4%Сам.
MAXIFE87.9%Сам.
MMMLU86.7%Сам.
MMMU83.9%Сам.
MMStar82.9%Сам.
Include82.8%Сам.
LiveCodeBench v678.9%Сам.
MMMU-Pro76.9%Сам.
IFBench76.1%Сам.
SimpleVQA0.62 / 100Сам.
LongBench v260.2%Сам.
NOVA-6358.6%Сам.

Grounding

RefCOCO-avg0.91 / 100Сам.
ScreenSpot Pro70.4%Сам.
RefSpatialBench0.69 / 100Сам.

Healthcare

VideoMMMU82.0%Сам.
SlakeVQA81.6%Сам.
MedXpertQA67.3%Сам.
PMC-VQA63.3%Сам.

Image To Text

OCRBench92.1%Сам.

Language

LingoQA80.8%Сам.
WMT24++78.3%Сам.

Long Context

MLVU87.3%Сам.
LVBench74.4%Сам.
AA-LCR66.9%Сам.
MMLongBench-Doc0.59 / 100Сам.

Math

HMMT 202591.4%Сам.
HMMT2590.3%Сам.
MathVista-Mini87.4%Сам.
MathVision86.2%Сам.
DynaMath85.9%Сам.
CodeForces0.85 / 3000Сам.
PolyMATH68.9%Сам.
Humanity's Last Exam47.5%Сам.

Multimodal

VLMsAreBlind96.7%Сам.
AI2D93.3%Сам.
V*93.2%Сам.
MMBench-V1.192.8%Сам.
OmniDocBench 1.589.8%Сам.
VideoMME w sub.87.3%Сам.
VideoMME w/o sub.83.9%Сам.
CC-OCR81.8%Сам.
CharXiv-R77.2%Сам.
MVBench76.6%Сам.
MMVU74.7%Сам.
BabyVision40.2%Сам.
ZEROBench-Sub0.36 / 100Сам.
Nuscene15.4%Сам.
ZEROBench0.09 / 100Сам.

Reasoning

CountBench0.97 / 100Сам.
BrowseComp-zh69.9%Сам.
Hallusion Bench67.6%Сам.
ERQA62.0%Сам.
Seal-044.1%Сам.
OJBench39.5%Сам.

Spatial Reasoning

RealWorldQA85.1%Сам.

Vision

ODinW44.5%Сам.

Индексы оценки AA

Intelligence Index
35.9
Coding Index
31.6
Tau2
0.8
Gpqa
0.8
Lcr
0.6
Ifbench
0.5
Scicode
0.4
Terminalbench Hard
0.3
Hle
0.1

Оценки категорий LLM Stats

Biology
90
Structured Output
80
Text-to-image
80
Video
80
Chemistry
80
Embodied
80
Finance
80
General
80
Grounding
80
Healthcare
80
Image To Text
80
Instruction Following
80
Language
80
Legal
80
Math
80
Physics
80
Spatial Reasoning
70
Vision
70
Economics
70
Frontend Development
70
Long Context
70
Multimodal
70
Reasoning
70
Tool Calling
60
Agents
60
Code
60
Communication
60
Search
60
Spatial
20
3d
20

Цены

Цена ввода$0.4 / 1M tokens
Цена вывода$3.2 / 1M tokens
Смешанная цена (3:1)$1.1 / 1M tokens

Скорость

Токенов/сек146.3 tokens/s
Задержка первого токена1.23s
Время до первого ответа1.23s

Доступные провайдеры

(Внутренние единицы LS)
ПровайдерЦена вводаЦена вывода
Novita400K3.2M

Внешние ссылки

LLM Stats