Skip to content

Benchmark

Arena ELO - Japanese

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

Models measured
245
Scores recorded
286
Sources
1
IQ weight
1.65

Score distribution

1,296.5
Mean
1,336.3
Median
150
Std dev
1,061.5
10th percentile
1,461.1
90th percentile
878.3
Min
878.31,557.4
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.65 for chat models, measured over 245 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
201Yi Lightning· LMArena data1,228.2
202DeepSeek V2.51210· LMArena data1,228.1
203GPT-5 nanoHigh· LMArena data1,226.7
204GLM 4 Plus0111· LMArena data1,225.5
205MiniMax M1· LMArena data1,217.3
206GPT-4o mini2024-07-18· LMArena data1,215.8
207Athene V2 Chat· LMArena data1,215.8
208Grok 2 mini2024-08-13· LMArena data1,213.1
209GPT-4 Turbo Preview1106· LMArena data1,209.5
210Mistral Small 3.1 24B Instruct2503· LMArena data1,208.4
211Llama 3.1 405B InstructBF16· LMArena data1,208.1
212GLM 4 Plus· LMArena data1,207.6
213Qwen2.5-Plus1127· LMArena data1,206.9
214Llama 4 Scout· LMArena data1,206.7
215Llama 4 Maverick· LMArena data1,206.2
216Qwen Max0919· LMArena data1,205.3
217Claude 3 Opus20240229· LMArena data1,204.2
218Nova Pro 1.0· LMArena data1,200.3
219GPT-4.1 nano2025-04-14· LMArena data1,198.6
220GPT-4 Turbo Preview0125· LMArena data1,195.4
221GPT-4 Turbo2024-04-09· LMArena data1,194.1
222DeepSeek V2.5· LMArena data1,190.3
223Mistral Large 22407· LMArena data1,187.7
224Gemini 1.5 Flash001· LMArena data1,186.5
225Qwen2.5 72B Instruct· LMArena data1,180
226Magistral Medium2506· LMArena data1,175.7
227Gemma 2 27B IT· LMArena data1,175.6
228Claude Haiku 3.520241022· LMArena data1,175.3
229MiniMax M2.5· LMArena data1,171.9
230Llama 3.1 405B InstructFP8· LMArena data1,171.6
231Mistral Large2411· LMArena data1,170.3
232Command R+08-2024· LMArena data1,166.4
233Aya Expanse 32B· LMArena data1,163.4
234Phi 4· LMArena data1,158.5
235Command R+· LMArena data1,154.9
236Nova Micro 1.0· LMArena data1,154.4
237Nova Lite 1.0· LMArena data1,153.6
238Llama 3.3 70B Instruct· LMArena data1,150.8
239Gemini 1.5 Flash-8B001· LMArena data1,150.3
240Gemma 2 9B IT· LMArena data1,144.5
241Gemma 2 9B IT SimPO· LMArena data1,139.9
242Athene 70B0725· LMArena data1,138.5
243GPT 40314· LMArena data1,136.7
244Llama 3.1 70B Instruct· LMArena data1,132.6
245Claude 3 Sonnet20240229· LMArena data1,131.4
246Nemotron-4 340B Instruct· LMArena data1,128.2
247DeepSeek Coder V2 Instruct· LMArena data1,126.6
248Aya Expanse 8B· LMArena data1,120.5
249Jamba 1.5 Large· LMArena data1,118.4
250GPT 40613· LMArena data1,114.3

Showing 201–250 of 286 results.

Data credits: LMArena, CC BY 4.0 (licence)