Qwen3 235B A22B 2507 Instruct
AlibabaQwenOpen WeightApache 2.0 · Usage Commercial
Description
Qwen3-235B-A22B-Instruct-2507 is the updated instruct version of Qwen3-235B-A22B featuring significant improvements in general capabilities including instruction following, logical reasoning, text comprehension, mathematics, science, coding and tool usage. It provides substantial gains in long-tail knowledge coverage across multiple languages and markedly better alignment with user preferences in subjective and open-ended tasks.
Date de sortie
2025-07-21
Paramètres
235.0B
Longueur du contexte
262K
Modalités
text
Radar de capacités
36
general
34
coding
76
reasoning
49
scienceest.
60
agents
0
multimodal
Science utilise un proxy de raisonnement lorsque les benchmarks scientifiques dédiés ne sont pas disponibles.
Classements
| Domaine | #Rang | Score | Source |
|---|---|---|---|
| Capacité agentique | 11 | 71.0 | LS |
| Classement codage | 231 | 39.0 | AA |
| Classement général | 228 | 46.0 | AA |
| Raisonnement mathématique | 96 | 78.0 | AA |
| Raisonnement | 80 | 55.0 | LS |
| Science | 171 | 52.0 | AA |
Scores de benchmarks (LLM Stats)
Agents
BFCL-v3
70.9%Aut.
Biology
GPQA
77.5%Aut.
Chemistry
SuperGPQA
62.6%Aut.
Code
Aider-Polyglot
57.3%Aut.
Communication
WritingBench
85.2%Aut.
Multi-IF
77.5%Aut.
Tau2 Retail
71.3%Aut.
Tau2 Airline
44.0%Aut.
Creativity
Creative Writing v3
87.5%Aut.
Arena-Hard v2
79.2%Aut.
Factuality
SimpleQA
54.3%Aut.
Finance
MMLU-Pro
83.0%Aut.
MMLU-ProX
79.4%Aut.
General
MMLU-Redux
93.1%Aut.
IFEval
88.7%Aut.
MultiPL-E
87.9%Aut.
CSimpleQA
84.3%Aut.
Include
79.5%Aut.
LiveBench 20241125
75.4%Aut.
LiveCodeBench v6
51.8%Aut.
Math
AIME 2025
70.3%Aut.
HMMT25
55.4%Aut.
PolyMATH
50.2%Aut.
Reasoning
ZebraLogic
95.0%Aut.
ARC-AGI
41.8%Aut.
Indices d'évaluation AA
Math Index71.7
Coding Index22.1
Intelligence Index18.2
Math 5001.0
Mmlu Pro0.8
Gpqa0.8
Aime0.7
Aime 250.7
Livecodebench0.5
Ifbench0.5
Scicode0.4
Tau20.3
Lcr0.3
Terminalbench Hard0.2
Hle0.1
Scores par catégorie LLM Stats
Structured Output80
Instruction Following80
Language80
Legal80
Finance80
Healthcare80
Biology80
Creativity80
Writing80
Math70
Physics70
Reasoning70
General70
Agents70
Chemistry70
Communication70
Code60
Economics60
Tool Calling60
Spatial Reasoning50
Multimodal50
Factuality50
Vision50
Tarification
Prix d'entrée$0.2 / 1M tokens
Prix de sortie$0.825 / 1M tokens
Prix mixte (3:1)$0.356 / 1M tokens
Vitesse
Tokens/sec67.2
Délai du premier token1.22s
Temps de réponse1.22s
Classement des Prix par Fournisseur
Classement des Prix par Fournisseur
23 fournisseurs
Moins cher: CortecsPlus cher: Scaleway
FournisseurEntréeSortie
1CortecsMoins cher
$0.062
$0.408
2SiliconFlow (China)
$0.09
$0.6
3NovitaAI
$0.09
$0.58
4Meganova
$0.09
$0.6
5SiliconFlow
$0.09
$0.6
6LLM Gateway
$0.09
$0.58
7Weights & Biases
$0.1
$0.1
8Chutes
$0.1
$0.6
9NanoGPT
$0.13
$0.5
10Abacus
$0.13
$0.6
11Jiekou.AI
$0.15
$0.8
12Venice AI
$0.15
$0.75
13AlibabaPRINCIPAL
$0.2
$0.825
14submodel
$0.2
$0.3
15Nebius Token Factory
$0.2
$0.6
16Synthetic
$0.2
$0.6
17Friendli
$0.2
$0.8
18Together AI
$0.2
$0.6
19Amazon Bedrock
$0.22
$0.88
20Vercel AI Gateway
$0.22
$0.88
21Vertex
$0.22
$0.88
22302.AI
$0.29
$1.143
23Scaleway
$0.75
$2.25
Comparer les prix entre différents fournisseurs API pour ce modèle.