Skip to content

Benchmark

Arena ELO - Industry Medicine and Healthcare

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

Models measured
336
Scores recorded
391
Sources
1
IQ weight
1.38

Score distribution

1,319.8
Mean
1,368.4
Median
152.6
Std dev
1,071.7
10th percentile
1,469.7
90th percentile
819.1
Min
819.11,538.2
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.38 for chat models, measured over 336 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
251Nemotron 3.5 Lightning 30B A3BNVFP4· LMArena data1,297.7
252O1 Mini· LMArena data1,296.8
253Athene V2 Chat· LMArena data1,296
254Granite 4.1 8B· LMArena data1,295.6
255DeepSeek V2.51210· LMArena data1,292
256Gemini 1.5 Flash002· LMArena data1,290
257GLM 4 Plus· LMArena data1,289.6
258Claude Sonnet 3.7Thinking 32K, 20250219· LMArena data1,288.2
259Olmo 3.1 32B Instruct· LMArena data1,287.1
260Llama 3.1 405B InstructBF16· LMArena data1,286.5
261GPT-4o mini2024-07-18· LMArena data1,281.8
262Reka Core20240904· LMArena data1,278.2
263GPT 4o2024-05-13· LMArena data1,278
264GPT-4.1 nano2025-04-14· LMArena data1,276.4
265Llama 4 Scout· LMArena data1,275.5
266Llama 4 Maverick· LMArena data1,275.5
267Llama 3.3 70B Instruct· LMArena data1,275.4
268Granite 4.2 3B· LMArena data1,272.5
269Athene 70B0725· LMArena data1,272.3
270Llama 3.1 405B InstructFP8· LMArena data1,270.6
271Mistral Small 3.1 24B Instruct2503· LMArena data1,266.3
272Olmo 3.1 32B Think· LMArena data1,260.7
273Nova Pro 1.0· LMArena data1,260.4
274DeepSeek V2.5· LMArena data1,259.3
275Qwen Max0919· LMArena data1,259.1
276Qwen2.5 72B Instruct· LMArena data1,258.5
277Claude Sonnet 3.720250219· LMArena data1,258.3
278Llama 3.1 70B Instruct· LMArena data1,257.2
279Claude 3.5 Sonnet20241022· LMArena data1,256.6
280Mistral Large2411· LMArena data1,254.7
281GPT 4o2024-08-06· LMArena data1,251.9
282Gemini Advanced0514· LMArena data1,249.4
283Magistral Medium2506· LMArena data1,246
284Mistral Large 22407· LMArena data1,245.6
285Claude 3.5 Sonnet20240620· LMArena data1,244.7
286GPT-4 Turbo2024-04-09· LMArena data1,239.4
287Command R+08-2024· LMArena data1,237.9
288GPT-4 Turbo Preview0125· LMArena data1,231.6
289Claude 3 Opus20240229· LMArena data1,231.2
290Hunyuan Large Vision· LMArena data1,229
291Claude Haiku 3.520241022· LMArena data1,227.6
292GPT-4 Turbo Preview1106· LMArena data1,224.3
293Aya Expanse 32B· LMArena data1,222.5
294Gemma 2 9B IT SimPO· LMArena data1,221.7
295Granite 4.0 H Small· LMArena data1,219.6
296Mistral Small 24B Instruct2501· LMArena data1,215.6
297Llama 3 70B Instruct· LMArena data1,214.1
298Gemini 1.5 Flash-8B001· LMArena data1,213.8
299Nova Lite 1.0· LMArena data1,212.6
300Gemini 1.5 Pro001· LMArena data1,212.2

Showing 251–300 of 391 results.

Data credits: LMArena, CC BY 4.0 (licence)