Skip to content

Benchmark

Arena ELO - Exclude Ties

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

Models measured
366
Scores recorded
423
Sources
1
IQ weight
1.28

Score distribution

1,247.7
Mean
1,316.3
Median
217.2
Std dev
908.5
10th percentile
1,465.2
90th percentile
533.7
Min
533.71,565.6
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.28 for chat models, measured over 366 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
101Mistral Large 3· LMArena data1,420.7
102DeepSeek R10528· LMArena data1,419.8
103Grok 3 Preview02-24· LMArena data1,418.6
104MiMo V2.5· LMArena data1,417.4
105MiniMax M3· LMArena data1,417.3
106GPT-6.1 TerraMax· LMArena data1,416.5
107LongCat-Flash-Chat Exp2602· LMArena data1,416.3
108Mistral Medium 3.12508· LMArena data1,415.3
109DeepSeek V3.2· LMArena data1,414.9
110Beam 1.0 501B A23B· LMArena data1,414.7
111DeepSeek-V4-Flash PreviewHigh· LMArena data1,414.6
112DeepSeek V3.2 ExpThinking· LMArena data1,414.2
113DeepSeek V3.2 Exp· LMArena data1,414
114GPT 5.1· LMArena data1,411.8
115MiMo V2 Omni· LMArena data1,411.8
116LongCat Flash Chat· LMArena data1,411.3
117GPT-4.5 Preview2025-02-27· LMArena data1,410.9
118Mistral Medium 3.5· LMArena data1,410
119Kimi K2.5Instant· LMArena data1,409.4
120GPT-5.5 Instant· LMArena data1,408.5
121nova-experimental-chat12-10· LMArena data1,408
122Qwen3 235B A22B Instruct 25072507· LMArena data1,407.7
123DeepSeek V3.2Thinking· LMArena data1,407.3
124Claude Opus 4.1Thinking 16K, 20250805· LMArena data1,407
125Qwen3 VL 235B· LMArena data1,406.9
126DeepSeek V3.1 TerminusThinking· LMArena data1,406.4
127DeepSeek V3.1· LMArena data1,405.8
128Claude Opus 4.120250805· LMArena data1,405.3
129Qwen3 Next 80B A3B Instruct· LMArena data1,404.5
130Gemini 2.5 Flash· LMArena data1,404.1
131Qwen3.5 122B-A10B· LMArena data1,403.2
132DeepSeek V3.1Thinking· LMArena data1,403.2
133GPT 5.2High· LMArena data1,403.1
134Gemini 3.1 Flash Lite· LMArena data1,402.5
135DeepSeek V3.1 Terminus· LMArena data1,402.1
136Qwen 3 235B A22B2507, thinking· LMArena data1,402.1
137nova-experimental-chat11-10· LMArena data1,401.1
138GPT-6.2 SolMax· LMArena data1,400.9
139GPT-6.2 AstraMax· LMArena data1,400.4
140Kimi K2 Thinking· LMArena data1,400.1
141Inkling Small· LMArena data1,399.7
142Mistral Large 4· LMArena data1,397.9
143GPT 5.2· LMArena data1,397.9
144Qwen3 Max2025-09-23· LMArena data1,397.7
145GPT-5.4 miniHigh· LMArena data1,397.6
146MiMo V2 FlashNon-Thinking· LMArena data1,396.6
147Grok 40709· LMArena data1,395.5
148NVIDIA Nemotron 3.5 Super VL 120BNVFP4· LMArena data1,392.9
149OpenAI o32025-04-16· LMArena data1,392.9
150Grok 4.1 FastReasoning· LMArena data1,391.9

Showing 101–150 of 423 results.

Data credits: LMArena, CC BY 4.0 (licence)