Skip to content

Benchmark

Arena ELO Vision - Creative Writing Vision

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

Models measured
93
Scores recorded
106
Sources
1
IQ weight
1.74

Score distribution

1,246.2
Mean
1,254
Median
53.6
Std dev
1,181.7
10th percentile
1,318.8
90th percentile
1,100
Min
1,1001,342.1
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.74 for chat models, measured over 93 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
51GPT-5.5 Instant· LMArena data1,255.8
52ChatGPT-4o Latest20250326· LMArena data1,255.2
53Gemini 3.5 Flash Lite· LMArena data1,254
54Step 5 Preview· LMArena data1,249.5
55Gemini 3.1 Flash Lite· LMArena data1,248
56GPT 5.1High· LMArena data1,246.8
57Claude Sonnet 5High· LMArena data1,245.9
58GPT-5.6 TerraxHigh· LMArena data1,245
59Gemini 2.5 Flash· LMArena data1,241.8
60Qwen 3.8 27B· LMArena data1,238.2
61MiMo V2.6 Flash· LMArena data1,237.5
62Kimi K2.5Instant· LMArena data1,236.1
63GPT-5.2 Chat20260210· LMArena data1,235.8
64Grok 4.1 FastReasoning· LMArena data1,234.4
65Grok 4.3· LMArena data1,234.2
66ERNIE-5.0-Preview1220· LMArena data1,232.7
67Grok 40709· LMArena data1,232.3
68MiniMax M3· LMArena data1,231.3
69MiMo V2.5· LMArena data1,230.1
70GPT 5.2· LMArena data1,226
71Qwen3.5 27B· LMArena data1,223
72GPT-5.4 miniHigh· LMArena data1,221.6
73Qwen3.5 122B-A10B· LMArena data1,220.7
74Qwen3 VL 235B· LMArena data1,220.6
75Mistral Large 3· LMArena data1,220.3
76GPT 5.2High· LMArena data1,215
77GPT-6.1 TerraMax· LMArena data1,214.9
78Mistral Large 4· LMArena data1,212.6
79GPT-5 Chat· LMArena data1,211.4
80GPT 4.12025-04-14· LMArena data1,211
81Grok 4.7xHigh· LMArena data1,203.9
82Qwen3 VL 235Bthinking· LMArena data1,203.5
83OpenAI o32025-04-16· LMArena data1,202
84Mistral Medium 3.12508· LMArena data1,200.8
85Inkling Small· LMArena data1,199.4
86Mistral Medium 3.5· LMArena data1,198.6
87GPT-6 LunaMax· LMArena data1,198.1
88Gemini 2.5 Flash-Lite PreviewNon-Thinking, 09-2025· LMArena data1,193.2
89NVIDIA Nemotron 3.5 Super VL 120BNVFP4· LMArena data1,192.9
90GPT-6 SolMax· LMArena data1,190.2
91GPT 5High· LMArena data1,184.3
92MiMo V2 Omni· LMArena data1,183.5
93GPT-4.1 mini2025-04-14· LMArena data1,183
94Mistral Small 3.22506· LMArena data1,182.6
95GPT-5 miniHigh· LMArena data1,181.7
96Gemini 2.5 Flash-Lite PreviewThinking, 06-17· LMArena data1,180.1
97Gemma 3 27B IT· LMArena data1,177.6
98Gemini 2.0 Flash001· LMArena data1,171.6
99o4 Mini2025-04-16· LMArena data1,167.9
100GPT 5.4 NanoHigh· LMArena data1,160

Showing 51–100 of 106 results.

Data credits: LMArena, CC BY 4.0 (licence)