Skip to content

Benchmark

Arena ELO - Industry Mathematical

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

Models measured
348
Scores recorded
402
Sources
1
IQ weight
1.25

Score distribution

1,321.4
Mean
1,367.7
Median
154.1
Std dev
1,094.5
10th percentile
1,481.4
90th percentile
833.6
Min
833.61,566.4
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.25 for chat models, measured over 348 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
101nova-experimental-chat12-10· LMArena data1,437.6
102LongCat-Flash-Chat Exp2602· LMArena data1,437.5
103Qwen3 VL 235Bthinking· LMArena data1,437.2
104DeepSeek V4 ProHigh, 20260813· LMArena data1,436.8
105Qwen 3 235B A22B2507, thinking· LMArena data1,436.3
106Grok 4 FastNon-Reasoning· LMArena data1,436.3
107Kimi K2 Thinking· LMArena data1,436.2
108Inkling Small· LMArena data1,436.2
109Grok 4.20 Beta 1· LMArena data1,434.9
110Grok 4.7xHigh· LMArena data1,434.8
111Qwen3 Next 80B A3B Instruct· LMArena data1,434.7
112DeepSeek V3.2 ExpThinking· LMArena data1,433.9
113Claude Opus 4.1Thinking 16K, 20250805· LMArena data1,433.6
114Qwen3.5 122B-A10B· LMArena data1,432.7
115Gemini 3 FlashMinimal· LMArena data1,432
116DeepSeek V4 Flash· LMArena data1,431.7
117Claude Haiku 5.5High· LMArena data1,431.7
118nova-experimental-chat11-10· LMArena data1,430.9
119GPT-5.2 Chat20260210· LMArena data1,430.8
120ERNIE-5.0-Preview1022· LMArena data1,430.7
121ERNIE 5.00110· LMArena data1,430.5
122Gemini 3.5 Flash Lite· LMArena data1,430.4
123Qwen3.5 27B· LMArena data1,430.4
124GLM 4.5· LMArena data1,430.2
125Claude Haiku 4.520251001· LMArena data1,429.6
126Doubao Seed 2.0 Pro· LMArena data1,429.4
127GPT-5.4 miniHigh· LMArena data1,428.1
128Mistral Large 3· LMArena data1,425.6
129Claude Opus 4.120250805· LMArena data1,425.6
130DeepSeek-V4-Flash PreviewHigh· LMArena data1,425
131Grok 40709· LMArena data1,424.7
132GPT-6 LunaMax· LMArena data1,424.2
133Solar Pro 4· LMArena data1,423.8
134OpenAI o32025-04-16· LMArena data1,422.8
135Gemini 2.5 Flash· LMArena data1,422.5
136Gemini 2.5 Flash Preview09-2025· LMArena data1,421
137GPT 5.1· LMArena data1,420.5
138Qwen3 32B· LMArena data1,419.9
139Qwen3 VL 235B· LMArena data1,419.6
140MiniMax-M2.1 Preview· LMArena data1,419.4
141Step 3.5 Flash· LMArena data1,419.4
142GPT 5.2· LMArena data1,418.2
143Gemini 3.1 Flash Lite· LMArena data1,416
144Qwen3 Max2025-09-23· LMArena data1,415.5
145DeepSeek V3.2 Exp· LMArena data1,415.5
146DeepSeek R10528· LMArena data1,415.5
147Grok 4.1Thinking· LMArena data1,415.5
148DeepSeek V3.1· LMArena data1,414.3
149Qwen3.5 Flash· LMArena data1,413.1
150nova-experimental-chat10-20· LMArena data1,412.8

Showing 101–150 of 402 results.

Data credits: LMArena, CC BY 4.0 (licence)