Skip to content

Benchmark

Arena ELO - Exclude Ties

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

Models measured
366
Scores recorded
423
Sources
1
IQ weight
1.28

Score distribution

1,247.7
Mean
1,316.3
Median
217.2
Std dev
908.5
10th percentile
1,465.2
90th percentile
533.7
Min
533.71,565.6
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.28 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
51Qwen3.7 Plus· LMArena data1,453.6
52GPT 5.4· LMArena data1,451.3
53Grok 4.20Reasoning, 0309, beta· LMArena data1,450.1
54DeepSeek V4 Pro· LMArena data1,450
55nova-experimental-chat26-02-10· LMArena data1,449.5
56Claude Opus 4.520251101· LMArena data1,449.3
57Grok 4.20 Multi-Agent0309, beta· LMArena data1,448.5
58MiMo V2.6 Flash· LMArena data1,448
59Doubao Seed 2.0 Pro· LMArena data1,447.4
60GPT-5.6 TerraxHigh· LMArena data1,446.7
61Claude Opus 4.5High, 32K, 20251101· LMArena data1,446.4
62Grok 4.5· LMArena data1,446
63GPT-6.1 SolMax· LMArena data1,444.5
64GLM 5· LMArena data1,443.3
65ERNIE 5.00110· LMArena data1,442.8
66Kimi K2.5Thinking· LMArena data1,441.6
67Qwen3.6 Max Preview· LMArena data1,441.5
68Gemma 4 31B IT· LMArena data1,440.9
69DeepSeek V4 ProHigh, 20260813· LMArena data1,440.8
70DeepSeek-V4-Pro PreviewHigh· LMArena data1,440.7
71Grok 4.20 Beta 1· LMArena data1,440.5
72Nemotron 3 UltraNVFP4· LMArena data1,440.3
73ERNIE-5.0-Preview1203· LMArena data1,439.9
74GPT 5.1High· LMArena data1,439.7
75Gemini 3 FlashMinimal· LMArena data1,438.9
76Inkling FP4· LMArena data1,438.7
77Step 5 Preview· LMArena data1,438.3
78Claude Sonnet 5High· LMArena data1,437.7
79GLM 4.6· LMArena data1,437.4
80GPT-6 AstraMax· LMArena data1,436.7
81Qwen3 Max Preview· LMArena data1,436.2
82Claude Sonnet 4.520250929· LMArena data1,434
83Hy3 (Deep Infra)· LMArena data1,433.3
84Qwen3.5 397B A17B· LMArena data1,433.2
85GPT-5.2 Chat20260210· LMArena data1,432.2
86Grok 4.1Thinking· LMArena data1,431.3
87Grok 4.1· LMArena data1,431
88GLM 4.7· LMArena data1,430.9
89Qwen 3.8 27B· LMArena data1,430.5
90MiMo V2 Pro· LMArena data1,430.4
91Qwen3.6 Plus· LMArena data1,430.1
92Gemini 3.5 Flash Lite· LMArena data1,429.1
93Gemma 4 26B A4B· LMArena data1,428.7
94Claude Sonnet 4.5High, 32K, 20250929· LMArena data1,426.9
95DeepSeek V4 Flash· LMArena data1,424.3
96ERNIE-5.0-Preview1022· LMArena data1,423.5
97GLM 4.5· LMArena data1,422.7
98GPT-5.6 LunaxHigh· LMArena data1,422.5
99Grok 4.6High· LMArena data1,421.3
100ChatGPT-4o Latest20250326· LMArena data1,421.2

Showing 51–100 of 423 results.

Data credits: LMArena, CC BY 4.0 (licence)