Skip to content

Benchmark

Arena ELO Text Factuality - Industry Medicine and Healthcare

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

Models measured
113
Scores recorded
128
Sources
1
IQ weight
2.07

Score distribution

1,475.5
Mean
1,474.9
Median
28.4
Std dev
1,444.2
10th percentile
1,511.8
90th percentile
1,365.9
Min
1,365.91,540.5
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 2.07 for chat models, measured over 113 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
101MiMo V2.5· LMArena data1,456.2
102GPT-5.4 miniHigh· LMArena data1,455.9
103Claude Opus 4.1Thinking 16K, 20250805· LMArena data1,455.4
104GPT-5.6 LunaxHigh· LMArena data1,455.2
105Gemini 2.5 Flash Preview09-2025· LMArena data1,454.2
106MiMo V2 FlashNon-Thinking· LMArena data1,450.7
107Qwen3.5 Flash· LMArena data1,450.3
108Grok 4.7xHigh· LMArena data1,450.1
109MiniMax M2.7· LMArena data1,449.7
110GPT-5.3 Chat· LMArena data1,449.6
111Inkling Small· LMArena data1,448.9
112Step 3.5 Flash· LMArena data1,447.9
113LongCat-Flash-Chat Exp2602· LMArena data1,447.5
114GPT-6 LunaMax· LMArena data1,446.4
115GPT-6 SolMax· LMArena data1,444.2
116Gemini 2.5 Flash-Lite PreviewNon-Thinking, 09-2025· LMArena data1,443.9
117Kimi K2 Thinking· LMArena data1,442.3
118Claude Haiku 5.5High· LMArena data1,440.1
119Grok 4.3· LMArena data1,437.7
120Claude Haiku 4.520251001· LMArena data1,434.2
121GPT 5.4 NanoHigh· LMArena data1,433.7
122Grok 4.1 FastReasoning· LMArena data1,430.2
123MiniMax-M2.1 Preview· LMArena data1,427.9
124Solar Pro 4· LMArena data1,424
125Trinity Large Thinking· LMArena data1,409.9
126MiniMax M2.5· LMArena data1,407.5
127Trinity Large Preview· LMArena data1,394.1
128Nemotron 3.5 Lightning 30B A3BNVFP4· LMArena data1,365.9

Showing 101–128 of 128 results.

Data credits: LMArena, CC BY 4.0 (licence)