Noometry

Model comparison

Grok 4.20 (Non-Reasoning) vs Qwen3.8 27B

Grok 4.20 (Non-Reasoning) is the stronger model overall, scoring 48.5 to 45.3 on the Noometry Index.

Last verified . 25 shared benchmarks.

Grok 4.20 (Non-Reasoning) xAI

48.5

Rank #48 Confirmed

Qwen3.8 27B Alibaba (Qwen)

45.3

Rank #64 Confirmed

Summary

  • They share 25 benchmarks with published results for both. Grok 4.20 (Non-Reasoning) scores higher in 5 categories and Qwen3.8 27B in 4 categories; 6 gaps are clear of the uncertainty.
  • The widest gap is in reasoning, where Grok 4.20 (Non-Reasoning) leads 52.2 to 31.5.
  • The biggest single-benchmark swing is NYT Connections (extended): 85.4% for Grok 4.20 (Non-Reasoning) and 54.5% for Qwen3.8 27B.
  • Qwen3.8 27B is cheaper at $0.99 / $1.49 per million input/output tokens, against $1.25 / $2.50 for Grok 4.20 (Non-Reasoning).
  • Grok 4.20 (Non-Reasoning) accepts more context: 1M tokens versus 262K.
  • Qwen3.8 27B has downloadable open weights; the other is API-only.

Side by side

Grok 4.20 (Non-Reasoning)Qwen3.8 27B
ProviderxAIAlibaba (Qwen)
Noometry Index48.545.3
Released2026-02-172026-08-14
WeightsProprietaryOpen
Context window1M262K
Max output30K33K
Input $ / M tokens$1.25$0.99
Output $ / M tokens$2.50$1.49
Results tracked4628

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

Category by category

Coding Qwen3.8 27B leads

Grok 4.20 (Non-Reasoning): 41.8 (#104), Qwen3.8 27B: 50.3 (#42)

BenchmarkGrok 4.20 (Non-Reasoning)Qwen3.8 27B
LMArena WebDev13741592
LMArena Coding14581487
SciCode—46.6%
WeirdML52.3%—
ALE-Bench1,150—

Agentic & Tool Use Not comparable

Grok 4.20 (Non-Reasoning): 34.3 (#43), Qwen3.8 27B: —

BenchmarkGrok 4.20 (Non-Reasoning)Qwen3.8 27B
Terminal-Bench57.3%—
τ²-bench Banking18%—
LMArena Search1189—
Vending-Bench 24,663—

Reasoning Grok 4.20 (Non-Reasoning) leads

Grok 4.20 (Non-Reasoning): 52.2 (#30), Qwen3.8 27B: 31.5 (#71)

BenchmarkGrok 4.20 (Non-Reasoning)Qwen3.8 27B
NYT Connections (extended)85.4%54.5%
LMArena Hard Prompts14511459
DTBench90.1%88%
LMCA38.7%41.4%
Epoch Capabilities Index152.01149.38
ARC-AGI-265.1%—
Kagi LLM Benchmark75%—
ARC-AGI-189.5%—
CritPt—5.4%
Chess Puzzles24%—
Thematic Generalization63.8%—
Surface Evolver Bench—45%
ForecastBench61.4—

Math Grok 4.20 (Non-Reasoning) leads

Grok 4.20 (Non-Reasoning): 48.0 (#54), Qwen3.8 27B: 35.7 (#139)

BenchmarkGrok 4.20 (Non-Reasoning)Qwen3.8 27B
ProofBench14%16%
LMArena Math14551453
FrontierMath (Tiers 1-3)44.9%—
FrontierMath Tier 417.1%—
OTIS Mock AIME 2024-202592.2%—

Knowledge Grok 4.20 (Non-Reasoning) leads

Grok 4.20 (Non-Reasoning): 52.8 (#54), Qwen3.8 27B: 40.9 (#105)

BenchmarkGrok 4.20 (Non-Reasoning)Qwen3.8 27B
LMArena Expert14401482
GPQA Diamond89.3%—
SimpleQA Verified30.2%—

Multimodal Qwen3.8 27B leads

Grok 4.20 (Non-Reasoning): 33.2 (#88), Qwen3.8 27B: 40.5 (#36)

BenchmarkGrok 4.20 (Non-Reasoning)Qwen3.8 27B
LMArena Vision12631270
Blueprint-Bench 20%—
LMArena Document1416—

Multilingual Too close to call

Grok 4.20 (Non-Reasoning): 54.7 (#38), Qwen3.8 27B: 54.0 (#54)

BenchmarkGrok 4.20 (Non-Reasoning)Qwen3.8 27B
LMArena Non-English14401431
LMArena Chinese14811504
LMArena French14771468
LMArena German14631435
LMArena Japanese14501392
LMArena Korean14171389
LMArena Russian14571421
LMArena Spanish14421444

Instruction Following Too close to call

Grok 4.20 (Non-Reasoning): 75.0 (#79), Qwen3.8 27B: 75.8 (#54)

BenchmarkGrok 4.20 (Non-Reasoning)Qwen3.8 27B
LMArena Instruction Following14201436

Long Context Grok 4.20 (Non-Reasoning) leads

Grok 4.20 (Non-Reasoning): 45.4 (#33), Qwen3.8 27B: 44.3 (#66)

BenchmarkGrok 4.20 (Non-Reasoning)Qwen3.8 27B
LMArena Longer Query14361448
CL-bench22.2%—
CL-bench Life11.9%—

Writing & Preference Too close to call

Grok 4.20 (Non-Reasoning): 65.9 (#42), Qwen3.8 27B: 65.9 (#41)

BenchmarkGrok 4.20 (Non-Reasoning)Qwen3.8 27B
LMArena Text14511441
LMArena Creative Writing14391382
EQ-Bench Creative Writing15741671
LMArena Multi-Turn14561443

Frequently asked questions

Is Grok 4.20 (Non-Reasoning) better than Qwen3.8 27B?

Grok 4.20 (Non-Reasoning) is the stronger model overall, scoring 48.5 to 45.3 on the Noometry Index.

Which is cheaper, Grok 4.20 (Non-Reasoning) or Qwen3.8 27B?

Qwen3.8 27B is cheaper. It lists at $0.99 per million input tokens and $1.49 per million output tokens; Grok 4.20 (Non-Reasoning) lists at $1.25 and $2.50.

Is Grok 4.20 (Non-Reasoning) or Qwen3.8 27B better for coding?

Qwen3.8 27B scores higher on coding benchmarks: 50.3 versus 41.8 in the Noometry coding category.

Which has the bigger context window?

Grok 4.20 (Non-Reasoning) does, with 1M tokens against 262K.

How many benchmarks do Grok 4.20 (Non-Reasoning) and Qwen3.8 27B share?

25 benchmarks have published results for both models. Grok 4.20 (Non-Reasoning) has 46 scored results on Noometry and Qwen3.8 27B has 28.

Related comparisons

Go deeper