The LLM benchmark aggregator.

Qwen3.5 4B

Alibaba · 4B

§ 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.

Agents50.3−17.8—/267✱0/5

Tool use50.7−23.3—/81✱0/1

Reasoning50.7−16.2—/178✱0/4

Mathematics53.6−12.1—/140✱0/2

Science52.6−10.9—/122✱0/1

Expert reasoning48.6−12.3—/0✱✱

Coding47.5−21.6—/164✱0/5

Agentic coding48.4−20.3—/157✱0/4

Code generation37.1−32.9—/77✱0/2

§ 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 Qwen3.5 4B, left for the other.

GPT-4owon 6 · lost 0

Gemma 4 E4Bwon 6 · lost 0

Mistral Large 3won 6 · lost 0

Claude 3.5 Haikuwon 6 · lost 0

Mistral Medium 3.5won 6 · lost 0

GPT-4.1 Nanowon 7 · lost 0

GPT-4o Miniwon 7 · lost 0

Llama 4 Maverick Instructwon 7 · lost 0

Granite 4.2 3Bwon 8 · lost 1

NVIDIA Nemotron 3 Nano 4Bwon 8 · lost 1

Qwen3.5 0.8Bwon 6 · lost 1

DeepSeek V3won 6 · lost 1

GPT-4.1 Miniwon 6 · lost 1

GPT-5.4 Nanowon 6 · lost 1

GPT OSS 120Bwon 5 · lost 1

Nemotron 3 Ultra 550B A55Bwon 5 · lost 1

Grok Build 0.1won 5 · lost 1

Granite 4.2 30Bwon 6 · lost 2

K2 Horizon 0.9Bwon 5 · lost 2

OpenBMB 1Bwon 5 · lost 2

Gemini 3.5 Flash Litewon 5 · lost 2

Grok 4.3won 5 · lost 2

Command Awon 5 · lost 2

Claude 3.5 Sonnetwon 4 · lost 2

Claude 3.7 Sonnetwon 4 · lost 2

Inkling Smallwon 4 · lost 2

Mimo V2.5won 4 · lost 2

Qwen3.6 27Bwon 4 · lost 2

G9v3 3Bwon 6 · lost 3

Inklingwon 6 · lost 3

MiniMax M3won 6 · lost 3

Nemotron 3 Ultrawon 5 · lost 3

Gemma 4 31Bwon 5 · lost 3

Nemotron 3 Superwon 5 · lost 3

Qwen3.6 35B A3Bwon 5 · lost 3

Qwen3.5 2Bwon 4 · lost 3

Gemini 3.1 Prowon 4 · lost 3

Claude Sonnet 4won 4 · lost 3

GPT-5.4 Miniwon 4 · lost 3

Qwen3.6 Pluswon 4 · lost 3

Claude Haiku 4.5won 4 · lost 3

Claude Opus 4won 3 · lost 3

DeepSeek R1won 3 · lost 3

Gemini 3.1 Flash Litewon 3 · lost 3

Muse Sparkwon 3 · lost 3

Mimo V2.5 Prowon 3 · lost 3

DeepSeek V4 Flash Visionwon 3 · lost 3

O3 Miniwon 3 · lost 4

Qwen3.7 Maxwon 3 · lost 4

Kimi K2.6won 3 · lost 4

GPT-4.1won 3 · lost 4

Kimi K2won 3 · lost 4

GLM-4.6won 3 · lost 4

Granite 4.2 8Bwon 3 · lost 5

G9v3 39A5Bwon 3 · lost 5

K2 Horizon 32Bwon 3 · lost 5

Muse Glimmer-30Bwon 3 · lost 5

K2 Horizon MoVA 36B A4Bwon 3 · lost 5

Gemini 2.5 Flashwon 3 · lost 5

Grok 3 Miniwon 2 · lost 4

MiniMax M2.7won 2 · lost 4

GPT-5.1won 2 · lost 4

Kimi K2.7 Codewon 2 · lost 4

Kimi K2.5won 2 · lost 4

Qwen3.8 Flash Nextwon 2 · lost 4

Grok 4.7won 2 · lost 4

Gemini 2.5 Prowon 2 · lost 5

Gemma 4 12Bwon 2 · lost 5

GPT-5 Miniwon 2 · lost 5

Grok 4.5won 2 · lost 5

Qwen3 235B A22B Instructwon 2 · lost 5

Claude Sonnet 4.5won 2 · lost 5

Grok 4.1 Fastwon 2 · lost 5

GLM-5.3 Flashwon 2 · lost 5

Gemini 3 Prowon 2 · lost 5

Qwen3.5 9Bwon 2 · lost 6

K2 Horizon 7Bwon 2 · lost 6

O4 Miniwon 2 · lost 6

DeepSeek V4 Flashwon 2 · lost 6

Kimi K3won 2 · lost 6

Claude Opus 4.5won 2 · lost 6

GLM-5.2won 2 · lost 7

Qwen3.8 27Bwon 2 · lost 7

GLM-4.7won 1 · lost 5

GPT-6 Lunawon 1 · lost 5

Muse Spark 1.2won 1 · lost 5

Gemini 3.5 Flashwon 1 · lost 6

Gemini 3.6 Flashwon 1 · lost 6

Muse Spark 1.1won 1 · lost 6

Claude Sonnet 4.6won 1 · lost 6

Gemini 3.8 Flashwon 1 · lost 6

Grok 4.6won 1 · lost 6

Qwen3.8 Maxwon 1 · lost 6

DeepSeek V4 Prowon 1 · lost 7

DeepSeek V4.1 Flashwon 1 · lost 7

GPT-5.6 Solwon 1 · lost 7

GLM-5.3won 1 · lost 7

K2 Horizon 3.7Bwon 1 · lost 8

GPT-5.6 Lunawon 1 · lost 8

Claude Opus 4.6won 0 · lost 6

GPT-5won 0 · lost 6

GLM-5.1won 0 · lost 6

GPT-6 Solwon 0 · lost 6

GPT-6 Astrawon 0 · lost 6

Muse Spark 1.3won 0 · lost 6

Claude Opus 5.5won 0 · lost 6

DeepSeek V3.2won 0 · lost 7

GPT-5.4won 0 · lost 7

Gemini 3.7 Flashwon 0 · lost 7

GPT-5.5won 0 · lost 7

Claude Opus 4.7won 0 · lost 7

Grok 4won 0 · lost 7

Claude Opus 4.8won 0 · lost 7

Claude Fable 5.1won 0 · lost 7

Claude Fable 5won 0 · lost 7

O3won 0 · lost 8

GPT-5.2won 0 · lost 8

K2 Horizon 375B A23Bwon 0 · lost 8

Claude Opus 5won 0 · lost 8

GPT-5.6 Terrawon 0 · lost 9

Claude Sonnet 5won 0 · lost 9

§ 3 · Sources

Where the numbers come from

1 publication, 8 figures. Every one links to the page it was read from.

✱✱ IFM — K2 Horizon launch ↗the model’s own publisher8 measuresread 2026-09-27

tau3-Banking 6.8%Terminal-Bench 2.1 25.8%SciCode 16.1%GPQA Diamond 77.1%SWE-bench Verified 41.2%HMMT Feb 2026 61.6%HLE 9.9%BFCL v4 55.7%

Badge

[![PublicAI Index](https://publicai.io/model-index/badge?model=qwen3-5-4b)](https://publicai.io/model-index/m/qwen3-5-4b)