Skip to content

Benchmark

Arena ELO - Multi Turn

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

Models measured
364
Scores recorded
421
Sources
1
IQ weight
1.21

Score distribution

1,304.7
Mean
1,350
Median
160.8
Std dev
1,057.9
10th percentile
1,467.2
90th percentile
740.3
Min
740.31,569.6
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.21 for chat models, measured over 364 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
101ERNIE 5.00110· LMArena data1,433.2
102GPT 5.1· LMArena data1,432.5
103Qwen3 235B A22B Instruct 25072507· LMArena data1,432.3
104GPT-6.1 TerraMax· LMArena data1,432.1
105DeepSeek-V4-Flash PreviewHigh· LMArena data1,431.7
106Grok 4.1Thinking· LMArena data1,430.3
107GPT-5.4 miniHigh· LMArena data1,429
108Mistral Large 3· LMArena data1,428.9
109Qwen3 VL 235B· LMArena data1,428
110Grok 4.6High· LMArena data1,427.4
111Kimi K2.5Instant· LMArena data1,427
112GLM 4.6· LMArena data1,426.8
113DeepSeek V3.2· LMArena data1,426.6
114GPT-5 Chat· LMArena data1,426.1
115GPT-6.2 AstraMax· LMArena data1,425.2
116Grok 3 Preview02-24· LMArena data1,424.1
117DeepSeek V3.2Thinking· LMArena data1,424.1
118Mistral Large 4· LMArena data1,423.3
119Mistral Medium 3.5· LMArena data1,423.1
120Hunyuan Vision 1.5 Thinking· LMArena data1,421.9
121DeepSeek V3.2 Exp· LMArena data1,421.2
122Beam 1.0 501B A23B· LMArena data1,421
123GPT-6.2 SolMax· LMArena data1,420
124GPT 5.2High· LMArena data1,419.5
125GPT 5.2· LMArena data1,418.4
126Mistral Medium 3.12508· LMArena data1,418.3
127Gemini 3.1 Flash Lite· LMArena data1,417.7
128ERNIE-5.0-Preview1022· LMArena data1,417.5
129LongCat-Flash-Chat Exp2602· LMArena data1,417.4
130NVIDIA Nemotron 3.5 Super VL 120BNVFP4· LMArena data1,416.8
131Inkling Small· LMArena data1,416.6
132Grok 40709· LMArena data1,416.2
133Qwen3.5 122B-A10B· LMArena data1,415.8
134Qwen3 Next 80B A3B Instruct· LMArena data1,415.6
135GLM 4.5· LMArena data1,415.3
136Kimi K2 Thinking· LMArena data1,414
137Grok 4 FastNon-Reasoning· LMArena data1,413.8
138GPT-6 SolMax· LMArena data1,413.4
139DeepSeek V3.2 ExpThinking· LMArena data1,412.8
140MiniMax M2.7· LMArena data1,412.4
141GPT-5.3 Chat· LMArena data1,412
142Nemotron 3 UltraNVFP4· LMArena data1,411
143DeepSeek V3.1 TerminusThinking· LMArena data1,410.5
144Qwen3.5 27B· LMArena data1,410
145Claude Haiku 4.520251001· LMArena data1,409.1
146ERNIE-5.0-Preview1203· LMArena data1,408.7
147nova-experimental-chat12-10· LMArena data1,408.1
148DeepSeek V3.1Thinking· LMArena data1,407.8
149Gemini 2.5 Flash· LMArena data1,407.7
150Qwen 3 235B A22B2507, thinking· LMArena data1,407

Showing 101–150 of 421 results.

Data credits: LMArena, CC BY 4.0 (licence)