Skip to content

Benchmark

Arena ELO Style-Controlled

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

Models measured
366
Scores recorded
423
Sources
1
IQ weight
1.10

Score distribution

1,341.3
Mean
1,363.7
Median
121
Std dev
1,152.6
10th percentile
1,474.5
90th percentile
953.4
Min
953.41,527
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.10 for chat models, measured over 366 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
51Grok 4.20Reasoning, 0309, beta· LMArena data1,472
52Claude Sonnet 4.6· LMArena data1,471.9
53Grok 4.20 Multi-Agent0309, beta· LMArena data1,471
54Claude Opus 4.520251101· LMArena data1,469.6
55GPT-6.2 AstraMax· LMArena data1,468.2
56GPT-5.6 TerraxHigh· LMArena data1,468
57ERNIE 5.1· LMArena data1,467.2
58MiMo V2.5 Pro· LMArena data1,466.3
59Grok 4.5· LMArena data1,465.7
60Grok 4.1Thinking· LMArena data1,465.4
61Qwen3.5-Max-Preview· LMArena data1,465.3
62GPT 5.4· LMArena data1,465.1
63GLM 5.1· LMArena data1,465.1
64GPT-6.2 SolMax· LMArena data1,464.7
65DeepSeek V4 ProHigh, 20260813· LMArena data1,463.8
66Kimi K2.6· LMArena data1,460.9
67Claude Sonnet 5High· LMArena data1,460.8
68Qwen3.6 Max Preview· LMArena data1,459.8
69Grok 4.1· LMArena data1,459.3
70Gemini 3 FlashMinimal· LMArena data1,458.2
71GPT-6 SolMax· LMArena data1,458.2
72DeepSeek V4 Pro· LMArena data1,458
73GLM 5· LMArena data1,457.4
74Grok 4.6High· LMArena data1,456.8
75Doubao Seed 2.0 Pro· LMArena data1,456.7
76Claude Sonnet 4.5High, 32K, 20250929· LMArena data1,456
77GPT 5.1High· LMArena data1,455.6
78Gemini 3.5 Flash Lite· LMArena data1,455.2
79Qwen3.7 Plus· LMArena data1,455.1
80Claude Sonnet 4.520250929· LMArena data1,455
81DeepSeek-V4-Pro PreviewHigh· LMArena data1,454.9
82Hy3 (Deep Infra)· LMArena data1,454.2
83GPT-5.6 LunaxHigh· LMArena data1,453.1
84Gemma 4 31B IT· LMArena data1,451.8
85GPT-5.3 Chat· LMArena data1,450.4
86Claude Opus 4.1Thinking 16K, 20250805· LMArena data1,449.8
87Kimi K2.5Thinking· LMArena data1,449.5
88Step 5 Preview· LMArena data1,448.8
89ERNIE-5.0-Preview1203· LMArena data1,448.5
90MiMo V2 Pro· LMArena data1,448.1
91Claude Opus 4.120250805· LMArena data1,447.9
92MiMo V2.6 Flash· LMArena data1,447.5
93GPT-5.4 miniHigh· LMArena data1,447.2
94ERNIE 5.00110· LMArena data1,445.8
95Gemini 2.5 Pro· LMArena data1,445.4
96GPT-4.5 Preview2025-02-27· LMArena data1,444.6
97Grok 4.7xHigh· LMArena data1,443.3
98Qwen3.6 Plus· LMArena data1,442.9
99ChatGPT-4o Latest20250326· LMArena data1,442.6
100GPT-6 LunaMax· LMArena data1,442.6

Showing 51–100 of 423 results.

Data credits: LMArena, CC BY 4.0 (licence)