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
301Claude Haiku 3.520241022· LMArena data1,171.2
302Nova Pro 1.0· LMArena data1,169.8
303Llama 3.1 Tulu 3 70B· LMArena data1,169.7
304Reka Core20240904· LMArena data1,156.4
305Granite 4.0 H Small· LMArena data1,154.2
306Gemini 1.5 Flash001· LMArena data1,142.4
307Jamba 1.5 Large· LMArena data1,139.4
308Mistral Small 24B Instruct2501· LMArena data1,129.1
309Gemma 2 27B IT· LMArena data1,128.6
310Qwen 2.5 Coder 32B Instruct· LMArena data1,126.1
311Command R+08-2024· LMArena data1,125.1
312GLM 40520· LMArena data1,123.6
313Gemma 2 9B IT SimPO· LMArena data1,122.4
314Llama 3.1 Nemotron 51B Instruct· LMArena data1,122.3
315Nemotron-4 340B Instruct· LMArena data1,122.2
316Nova Lite 1.0· LMArena data1,120.2
317Gemini 1.5 Flash-8B001· LMArena data1,117.8
318Llama 3 70B Instruct· LMArena data1,115.4
319Aya Expanse 32B· LMArena data1,114.8
320Claude 3 Sonnet20240229· LMArena data1,108.2
321OLMo 2 32B Instruct0325· LMArena data1,106.8
322Reka Flash20240904· LMArena data1,102.5
323Phi 4· LMArena data1,100.2
324Command R+· LMArena data1,090.8
325Gemma 2 9B IT· LMArena data1,090.1
326Qwen2 72B Instruct· LMArena data1,087.2
327GPT 40314· LMArena data1,085.8
328Nova Micro 1.0· LMArena data1,084.3
329Hunyuan Standard 256K· LMArena data1,079.3
330Claude 3 Haiku20240307· LMArena data1,075.1
331DeepSeek Coder V2 Instruct· LMArena data1,070.2
332Ministral 8B2410· LMArena data1,060.4
333Llama 3.1 Tulu 3 8B· LMArena data1,059.9
334GPT 40613· LMArena data1,058.3
335Jamba 1.5 Mini· LMArena data1,056.9
336Llama 3.1 8B Instruct· LMArena data1,055
337Aya Expanse 8B· LMArena data1,052.6
338Mistral Large2402· LMArena data1,046.7
339Qwen1.5-110B-Chat· LMArena data1,044.4
340Yi 1.5 34B Chat· LMArena data1,042.7
341Reka Flash 21BOnline, 20240226· LMArena data1,040.3
342Qwen1.5-72B-Chat· LMArena data1,032.3
343Reka Flash 21B20240226· LMArena data1,032
344Mistral Medium· LMArena data1,030.5
345Llama 3 8B Instruct· LMArena data1,030.1
346Mixtral 8x22B Instruct v0.1· LMArena data1,023.8
347QwQ 32B Preview· LMArena data1,022.5
348InternLM2.5 20B Chat· LMArena data1,010.1
349Gemini ProDev API· LMArena data1,008
350Gemma 2 2B IT· LMArena data1,004.3

Showing 301–350 of 423 results.

Data credits: LMArena, CC BY 4.0 (licence)