Skip to content

Benchmark

Arena ELO - Industry Software and It Services

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

Models measured
366
Scores recorded
423
Sources
1
IQ weight
1.29

Score distribution

1,322.6
Mean
1,374.8
Median
161.3
Std dev
1,081.5
10th percentile
1,486.1
90th percentile
719.3
Min
719.31,552.3
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.29 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
251GLM 4 Plus0111· LMArena data1,319.6
252Mercury· LMArena data1,318.6
253Qwen2.5-Plus1127· LMArena data1,316.1
254Claude 3.5 Sonnet20241022· LMArena data1,315.9
255Step-2 16K Exp202412· LMArena data1,315.1
256Claude Sonnet 3.720250219· LMArena data1,314.3
257Yi Lightning· LMArena data1,313.9
258DeepSeek V2.51210· LMArena data1,313.1
259Gemma 3 12B IT· LMArena data1,312.9
260GPT OSS 20b· LMArena data1,310.9
261Athene V2 Chat· LMArena data1,310
262Gemma 3n E4B IT· LMArena data1,307.5
263Hunyuan Large2025-02-10· LMArena data1,306.1
264GPT-4.1 nano2025-04-14· LMArena data1,304.8
265Gemini 1.5 Pro002· LMArena data1,303.5
266Mistral Small 3.1 24B Instruct2503· LMArena data1,301.4
267Olmo 3.1 32B Think· LMArena data1,300.7
268Llama 3.3 Nemotron Super 49B v1· LMArena data1,300.4
269Molmo 2 8B· LMArena data1,299.5
270Llama 4 Maverick· LMArena data1,298.7
271GPT 4o2024-05-13· LMArena data1,297.2
272DeepSeek V2.5· LMArena data1,295
273Grok 22024-08-13· LMArena data1,294.4
274Qwen2.5 72B Instruct· LMArena data1,292.1
275GLM 4 Plus· LMArena data1,291.1
276GPT-4o mini2024-07-18· LMArena data1,290.7
277Qwen Max0919· LMArena data1,290.6
278Llama 3.1 405B InstructBF16· LMArena data1,290.5
279Claude 3.5 Sonnet20240620· LMArena data1,288.3
280Llama 4 Scout· LMArena data1,285.5
281Hunyuan Standard2025-02-10· LMArena data1,284.6
282Magistral Medium2506· LMArena data1,283.8
283Llama 3.1 Nemotron 70B Instruct· LMArena data1,283.3
284Hunyuan Large Vision· LMArena data1,282
285Llama 3.1 405B InstructFP8· LMArena data1,280.8
286Gemini 1.5 Flash002· LMArena data1,278
287GPT 4o2024-08-06· LMArena data1,277.6
288Grok 2 mini2024-08-13· LMArena data1,277.4
289Mistral Large 22407· LMArena data1,276.1
290Mistral Large2411· LMArena data1,274
291Llama 3.3 70B Instruct· LMArena data1,272.8
292Claude Haiku 3.520241022· LMArena data1,272.3
293Gemini 1.5 Pro001· LMArena data1,272.2
294Nova Pro 1.0· LMArena data1,270.6
295Gemma 3 4B IT· LMArena data1,268.6
296GPT-4 Turbo2024-04-09· LMArena data1,267
297Gemini Advanced0514· LMArena data1,266.4
298Athene 70B0725· LMArena data1,265.2
299Qwen 2.5 Coder 32B Instruct· LMArena data1,260.9
300Llama 3.1 70B Instruct· LMArena data1,259.8

Showing 251–300 of 423 results.

Data credits: LMArena, CC BY 4.0 (licence)