ModelBeat
Models/Qwen3-235B-A22B (Jul 2025)
All models
A

Alibaba: Qwen3-235B-A22B (Jul 2025)

qwen3-235b-a22b-jul-2025

Qwen3-235B-A22B (Jul 2025) is Alibaba’s open-weights AI model, released Sep 1, 2025.

Context
Input
/ 1M
Output
/ 1M
Released
Sep 1, 2025
Modalities

Benchmarks

Scores on standardized evaluations. Higher is better — and rank shows where Qwen3-235B-A22B (Jul 2025) lands among all models tracked on Model Beat.

22.0
Intelligence Index
Epoch AI
22nd percentile of tracked models

Reasoning

5 evals
ARC-AGI11.0%

Abstract visual reasoning

ARC-AGI-21.3%

Harder abstract reasoning

GPQA Diamond80.1%

Graduate-level scientific reasoning

SimpleQA Verified50.1%

Factual accuracy & hallucination

WeirdML41.0%

Novel ML problem-solving

Math

3 evals
AIME 2024/202586.7%

Olympiad-qualifier math

FrontierMath8.5%

Research-level math problems

FrontierMath Tier 40.0%

Hardest research math

Compare Qwen3-235B-A22B (Jul 2025) with

Frequently asked questions

What is Qwen3-235B-A22B (Jul 2025)?

Qwen3-235B-A22B (Jul 2025) is a Qwen-family AI model, from Alibaba, released Sep 1, 2025.

Who created Qwen3-235B-A22B (Jul 2025)?

Qwen3-235B-A22B (Jul 2025) is an AI model developed by Alibaba, part of the Qwen family, released Sep 1, 2025.

How does Qwen3-235B-A22B (Jul 2025) perform on benchmarks?

On standardized evaluations tracked by Epoch AI, Qwen3-235B-A22B (Jul 2025) scores 50.1% on SimpleQA Verified, 86.7% on AIME 2024/2025, 80.1% on GPQA Diamond.

Is Qwen3-235B-A22B (Jul 2025) open source?

Qwen3-235B-A22B (Jul 2025) is released with open weights (Open weights (unrestricted)), so the model can be downloaded and self-hosted.

Benchmark data from Epoch AI (CC BY) and Artificial Analysis; pricing, specs & descriptions from OpenRouter. Data updated Jul 30, 2026.