Skip to content

Benchmark

Arena ELO

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

Models measured
366
Scores recorded
423
Sources
1
IQ weight
1.19

Score distribution

1,312
Mean
1,354.8
Median
146.6
Std dev
1,084.5
10th percentile
1,462.5
90th percentile
833.5
Min
833.51,539.3
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.19 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
51Claude Opus 4.8· LMArena data1,453.6
52GPT 5.4· LMArena data1,452.5
53DeepSeek V4 Pro· LMArena data1,451.3
54Grok 4.20Reasoning, 0309, beta· LMArena data1,451.3
55Grok 4.20 Multi-Agent0309, beta· LMArena data1,451
56Claude Opus 4.520251101· LMArena data1,450.9
57MiMo V2.6 Flash· LMArena data1,449.8
58GPT-5.6 TerraxHigh· LMArena data1,448.9
59nova-experimental-chat26-02-10· LMArena data1,448.7
60Doubao Seed 2.0 Pro· LMArena data1,448.6
61Claude Opus 4.5High, 32K, 20251101· LMArena data1,448.2
62Grok 4.5· LMArena data1,448
63GPT-6.1 SolMax· LMArena data1,446.5
64Qwen3.6 Max Preview· LMArena data1,446.4
65GLM 5· LMArena data1,446.1
66Nemotron 3 UltraNVFP4· LMArena data1,445.4
67DeepSeek-V4-Pro PreviewHigh· LMArena data1,445
68DeepSeek V4 ProHigh, 20260813· LMArena data1,444.7
69Kimi K2.5Thinking· LMArena data1,444.7
70Grok 4.20 Beta 1· LMArena data1,444.4
71ERNIE 5.00110· LMArena data1,444.3
72Step 5 Preview· LMArena data1,443.5
73Inkling FP4· LMArena data1,443
74GPT 5.1High· LMArena data1,442.6
75Gemma 4 31B IT· LMArena data1,442.5
76Gemini 3 FlashMinimal· LMArena data1,442.4
77ERNIE-5.0-Preview1203· LMArena data1,442.3
78Claude Sonnet 5High· LMArena data1,441.8
79GPT-6 AstraMax· LMArena data1,441
80GLM 4.6· LMArena data1,440.3
81Qwen3 Max Preview· LMArena data1,439.2
82Hy3 (Deep Infra)· LMArena data1,439
83Claude Sonnet 4.520250929· LMArena data1,438.5
84GPT-5.2 Chat20260210· LMArena data1,438.2
85Qwen3.5 397B A17B· LMArena data1,437.6
86Qwen 3.8 27B· LMArena data1,437.4
87Grok 4.1Thinking· LMArena data1,436.7
88MiMo V2 Pro· LMArena data1,436.5
89Grok 4.1· LMArena data1,436.3
90Qwen3.6 Plus· LMArena data1,436.2
91GLM 4.7· LMArena data1,435.3
92Gemini 3.5 Flash Lite· LMArena data1,435
93Claude Sonnet 4.5High, 32K, 20250929· LMArena data1,434.3
94Gemma 4 26B A4B· LMArena data1,434.2
95DeepSeek V4 Flash· LMArena data1,431.9
96GPT-5.6 LunaxHigh· LMArena data1,430.9
97GLM 4.5· LMArena data1,430.2
98Grok 4.6High· LMArena data1,430
99ERNIE-5.0-Preview1022· LMArena data1,429.6
100Mistral Large 3· LMArena data1,428.7

Showing 51–100 of 423 results.

Data credits: LMArena, CC BY 4.0 (licence)