Skip to content

Benchmark

Arena ELO Text Factuality - Industry Entertainment and Sports and Media

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

Models measured
150
Scores recorded
169
Sources
1
IQ weight
2.16

Score distribution

1,414.5
Mean
1,418.2
Median
43.7
Std dev
1,351.5
10th percentile
1,463.3
90th percentile
1,269.3
Min
1,269.31,504.5
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 2.16 for chat models, measured over 150 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-6.1 SolMax· LMArena data1,438.9
52GLM 5.3 Flash· LMArena data1,438.4
53GLM 5· LMArena data1,437.7
54Claude Opus 4.5High, 32K, 20251101· LMArena data1,437.2
55ERNIE 5.1· LMArena data1,437.2
56Kimi K2.6· LMArena data1,437.1
57DeepSeek V4 Pro· LMArena data1,437
58Muse Spark 1.1· LMArena data1,436.8
59GPT-5.6 TerraxHigh· LMArena data1,435.9
60Grok 4.6High· LMArena data1,435.4
61DeepSeek-V4-Pro PreviewHigh· LMArena data1,434
62Grok 4.20Reasoning, 0309, beta· LMArena data1,433.8
63GPT 5.1High· LMArena data1,433.5
64GPT-6.1 TerraMax· LMArena data1,432.9
65GPT-6.2 SolMax· LMArena data1,432.8
66Claude Sonnet 4.5High, 32K, 20250929· LMArena data1,432.8
67DeepSeek V4 ProHigh, 20260813· LMArena data1,431.8
68Grok 4.20 Multi-Agent0309, beta· LMArena data1,431.8
69Gemini 3.5 Flash Lite· LMArena data1,430.7
70Qwen3.7 Plus· LMArena data1,429.5
71Gemini 3 FlashMinimal· LMArena data1,429.4
72MiMo V2 Pro· LMArena data1,428.6
73GPT-6 LunaMax· LMArena data1,426.7
74GPT-5.6 LunaxHigh· LMArena data1,425.8
75Kimi K2.5Thinking· LMArena data1,425.7
76Grok 4.20 Beta 1· LMArena data1,425.7
77ChatGPT-4o Latest20250326· LMArena data1,424.9
78Step 5 Preview· LMArena data1,423.8
79DeepSeek V3.2 Exp· LMArena data1,422.4
80Claude Sonnet 5High· LMArena data1,421.5
81GPT-5.5 Instant· LMArena data1,421.5
82MiMo V2.6 Flash· LMArena data1,421.2
83Claude Opus 4.120250805· LMArena data1,421.1
84Grok 4.1Thinking· LMArena data1,419.3
85Hy3 (Deep Infra)· LMArena data1,418.2
86ERNIE 5.00110· LMArena data1,417.7
87GPT 5.1· LMArena data1,417.5
88Beam 1.0 501B A23B· LMArena data1,417.3
89Inkling FP4· LMArena data1,414.5
90GPT-5.2 Chat20260210· LMArena data1,414
91Qwen3.5 397B A17B· LMArena data1,413.6
92DeepSeek V4 Flash· LMArena data1,412.7
93MiMo V2.5· LMArena data1,412.6
94Grok 4.7xHigh· LMArena data1,411.9
95GLM 4.6· LMArena data1,411.3
96MiMo V2 Omni· LMArena data1,410.4
97GPT 5.2· LMArena data1,409.8
98ERNIE-5.0-Preview1203· LMArena data1,409.6
99DeepSeek V3.2· LMArena data1,409.6
100GLM 4.7· LMArena data1,409.6

Showing 51–100 of 169 results.

Data credits: LMArena, CC BY 4.0 (licence)