Skip to content

Benchmark

Arena ELO - Polish

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

Models measured
207
Scores recorded
242
Sources
1
IQ weight
2.07

Score distribution

1,386.8
Mean
1,405.7
Median
84.1
Std dev
1,270
10th percentile
1,481.7
90th percentile
1,072
Min
1,0721,546.7
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 207 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
101GLM 4.5· LMArena data1,422.1
102Claude Sonnet 4.520250929· LMArena data1,419.5
103Qwen3 VL 235B· LMArena data1,419.4
104DeepSeek V3.2Thinking· LMArena data1,419
105DeepSeek V3.2 Exp· LMArena data1,418.5
106GPT-5.4 miniHigh· LMArena data1,417.1
107DeepSeek-V4-Flash PreviewHigh· LMArena data1,416.1
108MiniMax M3· LMArena data1,414
109GPT-5.5 Instant· LMArena data1,413.3
110Gemini 2.5 Flash· LMArena data1,412.8
111Claude Opus 4.120250805· LMArena data1,412.6
112GPT 5High· LMArena data1,411.6
113MiMo V2.5· LMArena data1,410.4
114DeepSeek V3.1· LMArena data1,410.1
115Gemini 2.5 Flash Preview09-2025· LMArena data1,409.5
116DeepSeek V3.2· LMArena data1,408.9
117Claude Opus 4.1Thinking 16K, 20250805· LMArena data1,408.2
118MiMo V2 Omni· LMArena data1,407.1
119Qwen3 235B A22B Instruct 25072507· LMArena data1,406.8
120MiMo V2 FlashNon-Thinking· LMArena data1,406.1
121Step 5 Preview· LMArena data1,405.7
122DeepSeek V3.1Thinking· LMArena data1,405.3
123DeepSeek V3.1 Terminus· LMArena data1,405
124Grok 4.7xHigh· LMArena data1,404.9
125LongCat Flash Chat· LMArena data1,404.8
126GPT-5.6 LunaxHigh· LMArena data1,404.1
127Grok 4.1 FastReasoning· LMArena data1,400.3
128Qwen3.5 122B-A10B· LMArena data1,399.2
129nova-experimental-chat11-10· LMArena data1,398.9
130Grok 4.3· LMArena data1,398.3
131Qwen3 Next 80B A3B Instruct· LMArena data1,397.4
132GPT-5.3 Chat· LMArena data1,396.5
133Kimi K2 Thinking· LMArena data1,395.9
134Qwen3 Max2025-09-23· LMArena data1,393.7
135MiMo V2 FlashThinking· LMArena data1,392.9
136Qwen 3 235B A22B2507, thinking· LMArena data1,389.9
137GPT 4.12025-04-14· LMArena data1,388.6
138Step 3.5 Flash· LMArena data1,388.3
139GPT-6 SolMax· LMArena data1,388.3
140Hunyuan T120250711· LMArena data1,385.5
141Qwen 3 235B A22BNon-Thinking· LMArena data1,385.5
142MiniMax M2.7· LMArena data1,385.2
143Claude Haiku 5.5High· LMArena data1,384.7
144Qwen3.5 35B A3B· LMArena data1,383.6
145Nemotron 3 UltraNVFP4· LMArena data1,381.5
146GPT-6 LunaMax· LMArena data1,380.9
147nova-experimental-chat10-20· LMArena data1,379.9
148Grok 4 FastReasoning· LMArena data1,379.7
149Hunyuan TurboS20250416· LMArena data1,379.2
150GPT-5 miniHigh· LMArena data1,378.1

Showing 101–150 of 242 results.

Data credits: LMArena, CC BY 4.0 (licence)