Skip to content

Benchmark

Arena ELO - Creative Writing

People compare model answers and choose which they prefer. Higher scores are better.

Models measured
364
Scores recorded
421
Sources
1
IQ weight
1.14

Score distribution

1,287.4
Mean
1,318.1
Median
144.2
Std dev
1,079.9
10th percentile
1,442.3
90th percentile
793.9
Min
793.91,532.7
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.14 for chat models, measured over 364 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
201DeepSeek V3· LMArena data1,328.9
202nova-experimental-chat26-01-10· LMArena data1,327.8
203Step-2 16K Exp202412· LMArena data1,327.5
204GPT-5 miniHigh· LMArena data1,325.2
205Kimi K2 Preview0711· LMArena data1,323.9
206Step3· LMArena data1,322.2
207Trinity Large Preview· LMArena data1,319.3
208Command A03-2025· LMArena data1,319.1
209Gemini 2.0 Flash-Lite Preview02-05· LMArena data1,318.9
210O1 Preview· LMArena data1,318.1
211Qwen3 30B A3B Instruct2507· LMArena data1,317.3
212INTELLECT-3· LMArena data1,317.1
213Solar Pro 4· LMArena data1,317
214Claude Sonnet 3.720250219· LMArena data1,316.7
215nova-experimental-chat10-20· LMArena data1,316.4
216Qwen 3 235B A22B· LMArena data1,316.1
217Nemotron 3 Super 120B A12B· LMArena data1,315.2
218Qwen3 Next 80B A3Bthinking· LMArena data1,314.9
219Llama 3.1 Nemotron Ultra 253B v1· LMArena data1,314.2
220GPT 5.4 NanoHigh· LMArena data1,314.1
221GLM 4 Plus0111· LMArena data1,310.8
222Trinity Large Thinking· LMArena data1,307.9
223Step-1o Turbo202506· LMArena data1,307
224nova-experimental-chat10-09· LMArena data1,306.5
225Llama 3.3 Nemotron Super 49B v1.5· LMArena data1,305.7
226Mistral Small 3.22506· LMArena data1,304.3
227GPT-4.1 mini2025-04-14· LMArena data1,299.4
228MiniMax M1· LMArena data1,297.7
229Qwen3 32B· LMArena data1,297.3
230GLM 4.7 Flash· LMArena data1,295.6
231GLM 4.5V· LMArena data1,295.5
232o4 Mini2025-04-16· LMArena data1,294.2
233Qwen Plus0125· LMArena data1,292.4
234Claude 3.5 Sonnet20241022· LMArena data1,291.9
235Nova 2 Lite· LMArena data1,291.8
236GPT 4o2024-05-13· LMArena data1,291.4
237Ling Flash 2.0· LMArena data1,288.2
238Llama 3.3 Nemotron Super 49B v1· LMArena data1,287.9
239QwQ 32B· LMArena data1,287.5
240Gemma 3n E4B IT· LMArena data1,287
241MiniMax M2· LMArena data1,286.9
242Mercury 2· LMArena data1,286.8
243o3 MiniHigh· LMArena data1,285.7
244Gemini 1.5 Flash002· LMArena data1,285.3
245Granite 4.2 30B· LMArena data1,285.2
246DeepSeek V2.51210· LMArena data1,285
247Grok 22024-08-13· LMArena data1,284.2
248Gemini Advanced0514· LMArena data1,284.1
249Yi Lightning· LMArena data1,280
250GPT 4o2024-08-06· LMArena data1,274.5

Showing 201–250 of 421 results.

Data credits: LMArena, CC BY 4.0 (licence)