Saltar al contenido principal

Qwen3 235B A22B 2507 Instruct

AlibabaQwenOpen WeightApache 2.0 · Uso Comercial

Descripción

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.

Fecha de lanzamiento
2025-07-21
Parámetros
235.0B
Longitud del contexto
262K
Modalidades
text

Radar de capacidades

37
general
49
coding
76
reasoning
49
science
60
agents
0
multimodal

Rankings

Dominio#PosiciónPuntuaciónFuente
Capacidad agéntica53
52.0
LS
Ranking de codificación292
40.0
AA
Ranking general280
46.0
AA
Ciencia232
52.0
AA

Puntuaciones de benchmarks (LLM Stats)

(LLM Stats (zeroeval))

Agents

BFCL-v370.9%Aut.

Biology

GPQANYU + Cohere + Anthropic (2023)77.5%Aut.

Chemistry

SuperGPQA62.6%Aut.

Code

Aider-Polyglot57.3%Aut.

Communication

WritingBench85.2%Aut.
Multi-IF77.5%Aut.
Tau2 Retail71.3%Aut.
Tau2 Airline44.0%Aut.

Creativity

Creative Writing v387.5%Aut.
Arena-Hard v279.2%Aut.

Factuality

SimpleQA54.3%Aut.

Finance

MMLU-Pro83.0%Aut.
MMLU-ProX79.4%Aut.

General

MMLU-Redux93.1%Aut.
IFEvalGoogle Research (2023)88.7%Aut.
MultiPL-E87.9%Aut.
CSimpleQA84.3%Aut.
Include79.5%Aut.
LiveBench 2024112575.4%Aut.
LiveCodeBench v651.8%Aut.

Math

AIME 202570.3%Aut.
HMMT2555.4%Aut.
PolyMATH50.2%Aut.

Reasoning

ZebraLogic95.0%Aut.
ARC-AGI41.8%Aut.

Índices de evaluación AA

(Artificial Analysis)
Math Index(Artificial Analysis)
71.7
Intelligence Index(Artificial Analysis)
18.4
Math 500(OpenAI (2024), subset of Hendrycks et al. MATH (2021))
1.0
Mmlu Pro(TIGER-Lab (Univ. of Waterloo, Toronto, CMU, 2024))
0.8
Gpqa(NYU + Cohere + Anthropic (2023))
0.8
Aime(MAA (Mathematical Association of America))
0.7
Aime 25(MAA (Mathematical Association of America))
0.7
Livecodebench(UC Berkeley + MIT + Cornell (2024))
0.5
Ifbench(Google Research (2023))
0.5
Scicode(UIUC + Argonne National Lab (2024))
0.4
Tau2(Sierra + U Toronto + Vector Institute (2025))
0.3
Lcr(Artificial Analysis)
0.3
Terminalbench Hard(Stanford × Laude Institute (2026))
0.2
Hle(Center for AI Safety + Scale AI (2025))
0.1

Puntuaciones por categoría LLM Stats

(LLM Stats (zeroeval))
Legal
80
Structured Output
80
Instruction Following
80
Language
80
Finance
80
Healthcare
80
Biology
80
Creativity
80
Writing
80
Math
70
Physics
70
Reasoning
70
General
70
Agents
70
Chemistry
70
Communication
70
Code
60
Economics
60
Tool Calling
60
Multimodal
50
Spatial Reasoning
50
Factuality
50
Vision
50

Precios

Precio de entrada$0.7 / 1M tokens
Precio de salida$2.8 / 1M tokens
Precio mixto (3:1)$1.225 / 1M tokens

Velocidad

Tokens/seg0.0
Retraso del primer token0.00s
Tiempo hasta la respuesta0.00s

Ranking de Precios por Proveedor

Ranking de Precios por Proveedor

7 proveedores

Más barato: CortecsMás caro: GreenPT
ProveedorEntradaSalida
1CortecsMás barato
$0.069
$0.455
2LLM Gateway
$0.09
$0.58
3Venice AI
$0.15
$0.75
4302.AI
$0.29
$1.143
5AlibabaPRINCIPAL
$0.7
$2.8
6Scaleway
$0.75
$2.25
7GreenPT
$1.026
$3.078

Comparar precios entre diferentes proveedores de API para este modelo.

Fuentes externas