Benchmark · KnowledgeSettled

Lech Mazur Writing

Lech Mazur Writing · evaluates creative writing ability, assessing prose quality, narrative coherence, and stylistic sophistication.

Updated 2025-09-23
Models tested
46
Top score
87.1
Qwen3 Max
Median
77.0
min 61.7
Top-5 spread
σ 0.8
Settled

Best score over time · one chart, every benchmark

LECH MAZUR WRITING30 MODELS · FRONTIER RUNNING MAX0255075100SCORE ↑Nov 24Feb 25Apr 25Jul 25Sep 25RELEASE DATE →benchgecko.ai/benchmark/lech-mazur-writing · frontier
Frontier on Lech Mazur Writing rose from 81.8 to 87.1 in 10 months · +5.3 points · latest leader Qwen3 Max from Alibaba Qwen.
Pink dots = frontier records · 7 totalClick to open model page

46 models tested · sorted by score

#ModelScore
1Alibaba Qwen logoQwen3 Max87.1
2OpenAI logoGPT-586.0
3moonshotai logoKimi K2 071185.6
4DeepSeek logoDeepSeek V3.185.2
5Anthropic logoClaude Opus 4.184.7
6OpenAI logoo3 Pro84.4
7OpenAI logoo383.9
8Google DeepMind logoGemini 2.5 Pro83.8
9Anthropic logoClaude Opus 483.6
10OpenAI logoGPT-5 Mini83.1
11Alibaba Qwen logoQwen3 235B A22B83.0
12DeepSeek logoR183.0
13Alibaba Qwen logoQwen3 235B A22B Thinking 250782.4
14OpenAI logoGPT-4o (2024-11-20)81.8
15Anthropic logoClaude Sonnet 481.4
16xAI logoGrok 4 Fast81.1
17Anthropic logoClaude 3.7 Sonnet81.1
18Anthropic logoClaude 3.5 Sonnet80.3
19Alibaba Qwen logoQwQ 32B80.2
20Google DeepMind logoGemma 3 27B79.9
21Google DeepMind logoGemma 3 27B (free)79.9
22Mistral AI logoMistral Medium 377.3
23OpenAI logogpt-oss-120b77.1
24DeepSeek logoDeepSeek V377.0
25xAI logoGrok 476.9
26Google DeepMind logoGemini 2.5 Flash76.5
27xAI logoGrok 376.4
28OpenAI logoGPT-4.575.6
29Alibaba Qwen logoQwen3 30B A3B75.3
30OpenAI logoo4 Mini75.0
31Google DeepMind logoGemini 2.0 Flash Thinking (Jan 2025)73.8
32Anthropic logoClaude 3.5 Haiku73.5
33xAI logoGrok-3 mini73.5
34Alibaba Qwen logoQwen2.5-Max72.9
35Google DeepMind logoGemini 2.0 Flash71.5
36Google DeepMind logoGemini 2.0 Flash (Dec 2024)71.5
37OpenAI logoo170.2
38Mistral AI logoMistral Large69.0
39Mistral AI logoMistral Large 240769.0
40OpenAI logoGPT-4o-mini67.2
41OpenAI logoGPT-4o-mini (2024-07-18)67.2
42OpenAI logoo1-mini64.9
43xAI logoGrok-2 (Dec 2024)63.6
44Microsoft logoPhi 462.6
45Meta logoLlama 4 Maverick62.0
46OpenAI logoo3 Mini61.7

Same category · related evaluations