The LLM benchmark aggregator.

Qwen3 235B A22B Thinking

Alibaba · 235B · A22B · open weights

Strongest in Human preference (#124 of 341), weakest in Agents (#235 of 267). Above par in 2 of 4 scopes. Among the models it meets almost everywhere, it finishes behind Claude Opus 5.5 and Claude Fable 5 and ahead of O4 Mini and Claude 3.7 Sonnet.

§ 1 · Profile

What it is good at

Bars run from 50 — the average of the models each source lists — so right of the line is above par. The middle column is the gap to whoever leads that scope.

Human preference57−10.5#124/3411/1

Agents40.4−27.7#235/2672/5

Knowledge work37.5−36.1#142/1772/2

§ 2 · Head to head

What it beats, and what beats it

The same models turn up scope after scope. Counted once: where both were placed, who finished higher. Bars run right for Qwen3 235B A22B Thinking, left for the other.

Llama 3.2 1B Instructwon 3 · lost 0

Ministral 8Bwon 3 · lost 0

Llama 3.1 Nemotron Ultra 253B V1won 3 · lost 0

Llama 3.1 8B Instructwon 4 · lost 0

Granite 4.2 3Bwon 4 · lost 0

Llama 4 Scout Instructwon 4 · lost 0

Gemma 3 12B Itwon 4 · lost 0

O3 Miniwon 4 · lost 0

Gemma 3 27B Itwon 4 · lost 0

DeepSeek V3won 4 · lost 0

GPT-4won 3 · lost 1

GPT-4o Miniwon 3 · lost 1

Llama 3.3 70B Instructwon 3 · lost 1

GPT-4.1 Nanowon 3 · lost 1

Llama 4 Maverick Instructwon 3 · lost 1

Llama 3.2 3B Instructwon 2 · lost 1

Granite 3.1 8B Instructwon 2 · lost 1

Nova Microwon 2 · lost 1

Phi 4won 2 · lost 1

Nova Prowon 2 · lost 1

Gemma 3 4B Itwon 2 · lost 1

Mistral Largewon 2 · lost 1

GPT-5 Nanowon 2 · lost 1

Command Awon 2 · lost 1

Mistral Smallwon 2 · lost 1

Gemini 2.5 Flash Litewon 2 · lost 1

Qwen3 30B A3B Instructwon 2 · lost 1

Claude 3.7 Sonnetwon 2 · lost 1

O4 Miniwon 2 · lost 1

Granite 4.2 8Bwon 2 · lost 2

GPT OSS 20Bwon 2 · lost 2

Claude 3.5 Haikuwon 2 · lost 2

Nova 2 Litewon 2 · lost 2

Granite 4.2 30Bwon 2 · lost 2

Mercury 2won 2 · lost 2

Qwen3 32Bwon 2 · lost 2

GPT OSS 120Bwon 2 · lost 2

NVIDIA Nemotron 3 Super 120B A12Bwon 2 · lost 2

Trinity Largewon 2 · lost 2

Qwen3 Next 80B A3B Thinkingwon 2 · lost 2

GPT-4.1 Miniwon 2 · lost 2

Mimo V2 Flashwon 2 · lost 2

GPT-5 Miniwon 2 · lost 2

Qwen3.5 35B A3Bwon 2 · lost 2

DeepSeek R1won 2 · lost 2

Mistral Mediumwon 0 · lost 3

Gemini 2.5 Flashwon 0 · lost 3

Grok 4won 0 · lost 3

GPT-4.1won 0 · lost 3

Kimi K2won 0 · lost 3

Qwen3 235B A22B Instructwon 0 · lost 3

Grok 4.1 Fastwon 0 · lost 3

O3won 0 · lost 3

GPT-5.2won 0 · lost 3

Claude Opus 4.5won 0 · lost 3

Gemini 3 Prowon 0 · lost 3

GPT-5.4 Nanowon 0 · lost 4

Claude Sonnet 4won 0 · lost 4

Inkling Smallwon 0 · lost 4

Mistral Large 3won 0 · lost 4

Claude Haiku 4.5won 0 · lost 4

MiniMax M2.7won 0 · lost 4

Qwen3.5 122B A10Bwon 0 · lost 4

DeepSeek V3.1 Terminuswon 0 · lost 4

Muse Glimmerwon 0 · lost 4

GLM-4.6won 0 · lost 4

DeepSeek V3.2won 0 · lost 4

Mistral Medium 3.5won 0 · lost 4

Gemini 3.1 Flash Litewon 0 · lost 4

Mimo V2.5won 0 · lost 4

GPT-5won 0 · lost 4

Gemma 4 26B A4Bwon 0 · lost 4

Qwen3.8 27Bwon 0 · lost 4

Grok 4.7won 0 · lost 4

DeepSeek V4 Flashwon 0 · lost 4

MiniMax M3won 0 · lost 4

GLM-4.7won 0 · lost 4

GPT-6 Lunawon 0 · lost 4

Inklingwon 0 · lost 4

Grok 4.3won 0 · lost 4

Qwen3.5 397B A17Bwon 0 · lost 4

Qwen3.6 Pluswon 0 · lost 4

Gemini 2.5 Prowon 0 · lost 4

GPT-5.4 Miniwon 0 · lost 4

Kimi K2.5won 0 · lost 4

Gemma 4 31Bwon 0 · lost 4

Grok 4.6won 0 · lost 4

GPT-5.6 Lunawon 0 · lost 4

Gemini 3.5 Flash Litewon 0 · lost 4

GPT-5.1won 0 · lost 4

Qwen3.7 Pluswon 0 · lost 4

GPT-6 Solwon 0 · lost 4

Hy3won 0 · lost 4

Claude Sonnet 4.5won 0 · lost 4

Kimi K2.6won 0 · lost 4

Claude Sonnet 5won 0 · lost 4

DeepSeek V4 Prowon 0 · lost 4

GLM-5.1won 0 · lost 4

Grok 4.5won 0 · lost 4

GPT-5.6 Terrawon 0 · lost 4

Mimo V2.5 Prowon 0 · lost 4

Claude Sonnet 4.6won 0 · lost 4

GPT-5.5 Instantwon 0 · lost 4

GLM-5.3 Flashwon 0 · lost 4

Qwen3.7 Maxwon 0 · lost 4

GPT-5.4won 0 · lost 4

GLM-5.2won 0 · lost 4

DeepSeek V4.1 Flashwon 0 · lost 4

Gemini 3.5 Flashwon 0 · lost 4

GPT-6 Astrawon 0 · lost 4

Qwen3.8 Maxwon 0 · lost 4

Mimo V2.6 Prowon 0 · lost 4

GLM-5.3won 0 · lost 4

Claude Opus 4.8won 0 · lost 4

GPT-5.5won 0 · lost 4

Gemini 3.6 Flashwon 0 · lost 4

GPT-5.6 Solwon 0 · lost 4

Gemini 3.1 Prowon 0 · lost 4

Gemini 3.7 Flashwon 0 · lost 4

Kimi K3won 0 · lost 4

Claude Opus 5won 0 · lost 4

Muse Sparkwon 0 · lost 4

Muse Spark 1.1won 0 · lost 4

Gemini 3.8 Flashwon 0 · lost 4

Muse Spark 1.3won 0 · lost 4

Muse Spark 1.2won 0 · lost 4

Claude Fable 5.1won 0 · lost 4

Claude Opus 4.7won 0 · lost 4

Claude Fable 5won 0 · lost 4

Claude Opus 5.5won 0 · lost 4

§ 3 · Sources

Where the numbers come from

3 publications, 3 figures. Every one links to the page it was read from.

LMArena Text ↗1 measureread 2026-09-27

LMArena Text 1400

GDPval-AA ↗1 measureread 2026-09-27

GDPval-AA 302

AA-Briefcase ↗1 measureread 2026-09-27

AA-Briefcase 172

Badge

[![PublicAI Index](https://publicai.io/model-index/badge?model=qwen3-235b-a22b-thinking)](https://publicai.io/model-index/m/qwen3-235b-a22b-thinking)