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
151GPT-5.4 miniHigh· LMArena data1,369.6
152Qwen3.5 122B-A10B· LMArena data1,367.5
153Gemini 2.5 Flash-Lite PreviewThinking, 06-17· LMArena data1,367.2
154nova-experimental-chat26-02-10· LMArena data1,367.1
155Qwen3 VL 235B· LMArena data1,364.8
156GPT-5 Chat· LMArena data1,364.6
157GPT 5High· LMArena data1,364.4
158DeepSeek V30324· LMArena data1,364.2
159GPT 4.12025-04-14· LMArena data1,363.7
160Qwen3.5 27B· LMArena data1,362
161MiniMax-M2.1 Preview· LMArena data1,362
162GPT-6 LunaMax· LMArena data1,360.3
163Gemini 2.5 Flash-Lite PreviewNon-Thinking, 09-2025· LMArena data1,358.8
164OpenAI o32025-04-16· LMArena data1,358.7
165Hunyuan TurboS20250416· LMArena data1,358.6
166Qwen 3 235B A22BNon-Thinking· LMArena data1,356.6
167Step 3.5 Flash· LMArena data1,356.6
168GPT-5.3 Chat· LMArena data1,356.1
169DeepSeek R1· LMArena data1,354.1
170MiniMax M2.7· LMArena data1,353.9
171nova-experimental-chat11-10· LMArena data1,349.7
172nova-experimental-chat12-10· LMArena data1,349.5
173Kimi K2 Preview0905· LMArena data1,349.4
174MiMo V2 FlashThinking· LMArena data1,347.7
175o12024-12-17· LMArena data1,347.4
176Qwen3.5 35B A3B· LMArena data1,347
177Inkling 2 Small· LMArena data1,346.1
178Claude Sonnet 4Thinking 32K, 20250514· LMArena data1,345.5
179Gemma 3 27B IT· LMArena data1,345.4
180LongCat Flash Chat· LMArena data1,345.4
181GLM 4.6v· LMArena data1,345.1
182Qwen3 VL 235Bthinking· LMArena data1,344.1
183Mistral Medium 32505· LMArena data1,343.5
184GLM 4.5 Air· LMArena data1,343.3
185Qwen3.5 Flash· LMArena data1,343.2
186Claude Sonnet 420250514· LMArena data1,342.2
187Grok 3 Minibeta· LMArena data1,341.6
188Muse Glimmer· LMArena data1,341.4
189Mistral Large 4· LMArena data1,340.7
190Gemini 2.0 Flash001· LMArena data1,339.5
191Qwen2.5-Max· LMArena data1,338.7
192Inkling Small· LMArena data1,337.5
193Grok 3 MiniHigh· LMArena data1,334.4
194Claude Haiku 5.5High· LMArena data1,334.2
195Qwen3 Next 80B A3B Instruct· LMArena data1,333.2
196Qwen3 Coder· LMArena data1,333.1
197Gemini 1.5 Pro002· LMArena data1,332.6
198Claude Sonnet 3.7Thinking 32K, 20250219· LMArena data1,332
199Gemma 3 12B IT· LMArena data1,331.3
200MiniMax M2.5· LMArena data1,329.9

Showing 151–200 of 421 results.

Data credits: LMArena, CC BY 4.0 (licence)