Noometry

Model comparison

Grok-3 mini vs Qwen3.5-Flash

Qwen3.5-Flash is the stronger model overall, scoring 42.5 to 41.2 on the Noometry Index.

Last verified . 21 shared benchmarks.

Grok-3 mini xAI

41.2

Rank #141 Confirmed

Qwen3.5-Flash Alibaba (Qwen)

42.5

Rank #112 Confirmed

Summary

  • They share 21 benchmarks with published results for both. Grok-3 mini scores higher in 4 categories and Qwen3.5-Flash in 4 categories; 8 gaps are clear of the uncertainty.
  • The widest gap is in reasoning, where Qwen3.5-Flash leads 33.7 to 13.6.
  • The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 77.8% for Grok-3 mini and 84.4% for Qwen3.5-Flash.

Side by side

Grok-3 miniQwen3.5-Flash
ProviderxAIAlibaba (Qwen)
Noometry Index41.242.5
Released2025-04-092026-02-23
WeightsProprietaryProprietary
Context window—1M
Max output—66K
Input $ / M tokens—$0.10
Output $ / M tokens—$0.40
Results tracked3532

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

Category by category

Coding Grok-3 mini leads

Grok-3 mini: 40.8 (#131), Qwen3.5-Flash: 34.2 (#242)

BenchmarkGrok-3 miniQwen3.5-Flash
LMArena Coding13791412
Aider Polyglot49.3%—
LMArena WebDev—1244
WeirdML42.6%—
ALE-Bench—221.8

Agentic & Tool Use Not comparable

Grok-3 mini: —, Qwen3.5-Flash: —

BenchmarkGrok-3 miniQwen3.5-Flash
Vending-Bench 2—462.69

Reasoning Qwen3.5-Flash leads

Grok-3 mini: 13.6 (#334), Qwen3.5-Flash: 33.7 (#72)

BenchmarkGrok-3 miniQwen3.5-Flash
LMArena Hard Prompts13751403
Epoch Capabilities Index140.35143.98
ARC-AGI-20.4%—
Kagi LLM Benchmark61.3%—
ARC-AGI-116.5%—
Chess Puzzles—21%
Mystery Game Puzzles—20%
DTBench—82.9%
LMCA—29.1%

Math Grok-3 mini leads

Grok-3 mini: 42.1 (#85), Qwen3.5-Flash: 37.4 (#158)

BenchmarkGrok-3 miniQwen3.5-Flash
OTIS Mock AIME 2024-202577.8%84.4%
LMArena Math13861407
FrontierMath (Feb 2025 set)5.9%6.2%
FrontierMath (Tiers 1-3)—18.2%
Omni-MATH31.8%—
MATH Level 590.9%—
FrontierMath Tier 4 (v1)—0%

Knowledge Grok-3 mini leads

Grok-3 mini: 46.4 (#81), Qwen3.5-Flash: 43.2 (#93)

BenchmarkGrok-3 miniQwen3.5-Flash
GPQA Diamond76.3%82.3%
LMArena Expert13951407
SimpleQA Verified—20.3%
MMLU-Pro79.9%—
Confabulations10.8%—
Vectara Hallucination Rate—10.5%
GPQA (HELM)67.5%—

Multilingual Qwen3.5-Flash leads

Grok-3 mini: 48.1 (#145), Qwen3.5-Flash: 50.5 (#121)

BenchmarkGrok-3 miniQwen3.5-Flash
LMArena Non-English13521385
LMArena Chinese13871446
LMArena French13571412
LMArena German13491390
LMArena Japanese13421368
LMArena Korean13351344
LMArena Russian13531379
LMArena Spanish13811400

Instruction Following Grok-3 mini leads

Grok-3 mini: 78.5 (#9), Qwen3.5-Flash: 72.6 (#139)

BenchmarkGrok-3 miniQwen3.5-Flash
LMArena Instruction Following13571374
IFEval95.1%—

Long Context Qwen3.5-Flash leads

Grok-3 mini: 41.0 (#147), Qwen3.5-Flash: 42.4 (#124)

BenchmarkGrok-3 miniQwen3.5-Flash
LMArena Longer Query13721392
Fiction.LiveBench66.7%—

Writing & Preference Qwen3.5-Flash leads

Grok-3 mini: 52.5 (#169), Qwen3.5-Flash: 57.9 (#122)

BenchmarkGrok-3 miniQwen3.5-Flash
LMArena Text13701397
LMArena Creative Writing13421343
LMArena Multi-Turn13551393
Short-Story Creative Writing73.5%—
WildBench65.1%—

Frequently asked questions

Is Grok-3 mini better than Qwen3.5-Flash?

Qwen3.5-Flash is the stronger model overall, scoring 42.5 to 41.2 on the Noometry Index.

Is Grok-3 mini or Qwen3.5-Flash better for coding?

Grok-3 mini scores higher on coding benchmarks: 40.8 versus 34.2 in the Noometry coding category.

How many benchmarks do Grok-3 mini and Qwen3.5-Flash share?

21 benchmarks have published results for both models. Grok-3 mini has 35 scored results on Noometry and Qwen3.5-Flash has 32.

Related comparisons

Go deeper