Noometry

Model comparison

Grok-3 mini vs Qwen3.6 Max Preview

Qwen3.6 Max Preview is the stronger model overall, scoring 51.5 to 41.2 on the Noometry Index.

Last verified . 18 shared benchmarks.

Grok-3 mini xAI

41.2

Rank #141 Confirmed

Qwen3.6 Max Preview Alibaba (Qwen)

51.5

Rank #43 Confirmed

Summary

  • They share 18 benchmarks with published results for both. Grok-3 mini scores higher in 1 category and Qwen3.6 Max Preview in 7 categories; 8 gaps are clear of the uncertainty.
  • The widest gap is in reasoning, where Qwen3.6 Max Preview leads 41.7 to 13.6.
  • The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 77.8% for Grok-3 mini and 91.1% for Qwen3.6 Max Preview.

Side by side

Grok-3 miniQwen3.6 Max Preview
ProviderxAIAlibaba (Qwen)
Noometry Index41.251.5
Released2025-04-092026-04-20
WeightsProprietaryProprietary
Context window—262K
Max output—66K
Input $ / M tokens—$1.30
Output $ / M tokens—$7.80
Results tracked3529

Sponsored placements are available on pages like this one. Advertise on Noometry

Category by category

Coding Qwen3.6 Max Preview leads

Grok-3 mini: 40.8 (#131), Qwen3.6 Max Preview: 48.7 (#54)

BenchmarkGrok-3 miniQwen3.6 Max Preview
LMArena Coding13791471
SWE-bench Verified—76.7%
Aider Polyglot49.3%—
LMArena WebDev—1482
WeirdML42.6%—

Agentic & Tool Use Not comparable

Grok-3 mini: —, Qwen3.6 Max Preview: —

BenchmarkGrok-3 miniQwen3.6 Max Preview
Vending-Bench 2—4,254

Reasoning Qwen3.6 Max Preview leads

Grok-3 mini: 13.6 (#334), Qwen3.6 Max Preview: 41.7 (#53)

BenchmarkGrok-3 miniQwen3.6 Max Preview
LMArena Hard Prompts13751457
Epoch Capabilities Index140.35149.24
ARC-AGI-20.4%—
SimpleBench—63%
Kagi LLM Benchmark61.3%—
NYT Connections (extended)—74.1%
ARC-AGI-116.5%—
Chess Puzzles—20%
Mystery Game Puzzles—19%
DTBench—87.2%
LMCA—42.5%

Math Qwen3.6 Max Preview leads

Grok-3 mini: 42.1 (#85), Qwen3.6 Max Preview: 54.1 (#46)

BenchmarkGrok-3 miniQwen3.6 Max Preview
OTIS Mock AIME 2024-202577.8%91.1%
LMArena Math13861465
FrontierMath (Feb 2025 set)5.9%23.1%
Omni-MATH31.8%—
MATH Level 590.9%—
FrontierMath Tier 4 (v1)—4.2%

Knowledge Qwen3.6 Max Preview leads

Grok-3 mini: 46.4 (#81), Qwen3.6 Max Preview: 57.6 (#39)

BenchmarkGrok-3 miniQwen3.6 Max Preview
GPQA Diamond76.3%87.4%
LMArena Expert13951478
SimpleQA Verified—52%
MMLU-Pro79.9%—
Confabulations10.8%—
GPQA (HELM)67.5%—

Multilingual Qwen3.6 Max Preview leads

Grok-3 mini: 48.1 (#145), Qwen3.6 Max Preview: 54.2 (#48)

BenchmarkGrok-3 miniQwen3.6 Max Preview
LMArena Non-English13521437
LMArena Chinese13871487
LMArena French13571449
LMArena Russian13531445
LMArena Spanish13811454
LMArena German1349—
LMArena Japanese1342—
LMArena Korean1335—

Instruction Following Grok-3 mini leads

Grok-3 mini: 78.5 (#9), Qwen3.6 Max Preview: 75.7 (#55)

BenchmarkGrok-3 miniQwen3.6 Max Preview
LMArena Instruction Following13571438
IFEval95.1%—

Long Context Qwen3.6 Max Preview leads

Grok-3 mini: 41.0 (#147), Qwen3.6 Max Preview: 44.6 (#61)

BenchmarkGrok-3 miniQwen3.6 Max Preview
LMArena Longer Query13721457
Fiction.LiveBench66.7%—

Writing & Preference Qwen3.6 Max Preview leads

Grok-3 mini: 52.5 (#169), Qwen3.6 Max Preview: 63.8 (#60)

BenchmarkGrok-3 miniQwen3.6 Max Preview
LMArena Text13701447
LMArena Creative Writing13421435
LMArena Multi-Turn13551456
Short-Story Creative Writing73.5%—
WildBench65.1%—

Frequently asked questions

Is Grok-3 mini better than Qwen3.6 Max Preview?

Qwen3.6 Max Preview is the stronger model overall, scoring 51.5 to 41.2 on the Noometry Index.

Is Grok-3 mini or Qwen3.6 Max Preview better for coding?

Qwen3.6 Max Preview scores higher on coding benchmarks: 48.7 versus 40.8 in the Noometry coding category.

How many benchmarks do Grok-3 mini and Qwen3.6 Max Preview share?

18 benchmarks have published results for both models. Grok-3 mini has 35 scored results on Noometry and Qwen3.6 Max Preview has 29.

Related comparisons

Go deeper