Skip to content

Benchmark

Arena ELO - Longer Query

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

Models measured
345
Scores recorded
401
Sources
1
IQ weight
1.34

Score distribution

1,323.7
Mean
1,357.1
Median
137.1
Std dev
1,102.6
10th percentile
1,475.3
90th percentile
951.6
Min
951.61,549
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.34 for chat models, measured over 345 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 4.5· LMArena data1,462.2
52Claude Sonnet 5High· LMArena data1,462.2
53ERNIE 5.1· LMArena data1,462
54Muse Spark 1.1· LMArena data1,461.7
55GPT-6.1 SolMax· LMArena data1,460.3
56GPT-6 AstraMax· LMArena data1,458.1
57Qwen3.6 Max Preview· LMArena data1,458
58Grok 4.6High· LMArena data1,457.7
59DeepSeek V4 Pro· LMArena data1,457.6
60DeepSeek V4 ProHigh, 20260813· LMArena data1,456.9
61MiMo V2 Pro· LMArena data1,455.6
62Claude Opus 4.1Thinking 16K, 20250805· LMArena data1,454.7
63MiMo V2.6 Flash· LMArena data1,454.5
64Step 5 Preview· LMArena data1,454.1
65GPT-5.6 TerraxHigh· LMArena data1,453.4
66Qwen3.7 Plus· LMArena data1,453.4
67Gemini 3 Flash· LMArena data1,451.7
68Muse Spark· LMArena data1,450.2
69Gemini 2.5 Pro· LMArena data1,449.4
70GPT 5.1High· LMArena data1,447.3
71GLM 5· LMArena data1,446.2
72DeepSeek-V4-Pro PreviewHigh· LMArena data1,445.9
73Qwen 3.8 27B· LMArena data1,445.8
74GPT-6.1 TerraMax· LMArena data1,445.4
75MiMo V2.5· LMArena data1,445
76Gemma 4 31B IT· LMArena data1,445
77Claude Opus 4.120250805· LMArena data1,443.9
78Kimi K2.5Thinking· LMArena data1,443.8
79Hy3 (Deep Infra)· LMArena data1,443.7
80nova-experimental-chat26-02-10· LMArena data1,442
81Qwen3.5 397B A17B· LMArena data1,441.8
82MiMo V2 Omni· LMArena data1,440.4
83Qwen3.6 Plus· LMArena data1,439.9
84Grok 3 Preview02-24· LMArena data1,438.9
85Qwen3 Max Preview· LMArena data1,438.3
86MiniMax M3· LMArena data1,438.1
87Nemotron 3 UltraNVFP4· LMArena data1,437.2
88Inkling FP4· LMArena data1,437.1
89Grok 4.20Reasoning, 0309, beta· LMArena data1,436.7
90GPT-5.6 LunaxHigh· LMArena data1,436.3
91Kimi K2.5Instant· LMArena data1,435.5
92Gemini 3.5 Flash Lite· LMArena data1,435
93DeepSeek V4 Flash· LMArena data1,434.7
94GLM 4.7· LMArena data1,432.3
95GPT-6.2 AstraMax· LMArena data1,431.3
96Grok 4.20 Multi-Agent0309, beta· LMArena data1,431.3
97GPT 5.1· LMArena data1,429.3
98Gemini 3 FlashMinimal· LMArena data1,429.1
99DeepSeek V3.2· LMArena data1,428.4
100Doubao Seed 2.0 Pro· LMArena data1,428.1

Showing 51–100 of 401 results.

Data credits: LMArena, CC BY 4.0 (licence)