Skip to content

Benchmark

Arena ELO Text Factuality - Multi Turn

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

Models measured
142
Scores recorded
161
Sources
1
IQ weight
2.11

Score distribution

1,450.4
Mean
1,457.6
Median
36.7
Std dev
1,398.3
10th percentile
1,490.7
90th percentile
1,341.4
Min
1,341.41,524
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 2.11 for chat models, measured over 142 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
51Gemini 3 Flash· LMArena data1,471.5
52GLM 5.3 Flash· LMArena data1,470.3
53Claude Sonnet 4.6· LMArena data1,469.8
54GPT-5.2 Chat20260210· LMArena data1,469.2
55GPT 5.1High· LMArena data1,468.4
56GPT-6 LunaMax· LMArena data1,468.1
57ChatGPT-4o Latest20250326· LMArena data1,467.7
58Qwen3 Max Preview· LMArena data1,467.6
59Claude Sonnet 4.5High, 32K, 20250929· LMArena data1,466.7
60GPT-5.6 LunaxHigh· LMArena data1,465.2
61Grok 4.20Reasoning, 0309, beta· LMArena data1,464.9
62DeepSeek V4 ProHigh, 20260813· LMArena data1,464.5
63Claude Sonnet 5High· LMArena data1,464.4
64GPT-6.1 TerraMax· LMArena data1,464.3
65ERNIE 5.1· LMArena data1,464.3
66Gemini 2.5 Pro· LMArena data1,464.3
67Grok 4.20 Multi-Agent0309, beta· LMArena data1,464
68Gemini 3.5 Flash Lite· LMArena data1,464
69Gemini 3 FlashMinimal· LMArena data1,463.5
70Claude Opus 4.1Thinking 16K, 20250805· LMArena data1,461.6
71Kimi K2.6· LMArena data1,461.4
72Qwen3.7 Plus· LMArena data1,460.7
73GLM 5· LMArena data1,460.6
74GPT-5.4 miniHigh· LMArena data1,458.8
75GPT 5.1· LMArena data1,458.6
76MiMo V2.5· LMArena data1,458.5
77Grok 4.20 Beta 1· LMArena data1,458.4
78Inkling FP4· LMArena data1,458.3
79MiMo V2 Pro· LMArena data1,458.3
80Step 5 Preview· LMArena data1,457.8
81GPT-5.5 Instant· LMArena data1,457.6
82Claude Opus 4.120250805· LMArena data1,457.3
83Qwen3.5 397B A17B· LMArena data1,457.2
84Hy3 (Deep Infra)· LMArena data1,456.3
85MiMo V2.6 Flash· LMArena data1,455.4
86NVIDIA Nemotron 3.5 Super VL 120BNVFP4· LMArena data1,454.5
87GLM 4.7· LMArena data1,454.4
88DeepSeek V4 Flash· LMArena data1,453.9
89DeepSeek-V4-Pro PreviewHigh· LMArena data1,453.2
90Grok 4.6High· LMArena data1,452.4
91GPT 5.2· LMArena data1,452.3
92Beam 1.0 501B A23B· LMArena data1,451.6
93MiMo V2 Omni· LMArena data1,450.4
94GPT 5.2High· LMArena data1,450
95Kimi K2.5Thinking· LMArena data1,450
96GPT-6 SolMax· LMArena data1,449.8
97GPT-5 Chat· LMArena data1,449.1
98MiniMax M3· LMArena data1,447.4
99Qwen3 235B A22B Instruct 25072507· LMArena data1,447.4
100Qwen 3.8 27B· LMArena data1,446.5

Showing 51–100 of 161 results.

Data credits: LMArena, CC BY 4.0 (licence)