Skip to content

Benchmark

Arena ELO - Industry Mathematical

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

Models measured
348
Scores recorded
402
Sources
1
IQ weight
1.25

Score distribution

1,321.4
Mean
1,367.7
Median
154.1
Std dev
1,094.5
10th percentile
1,481.4
90th percentile
833.6
Min
833.61,566.4
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.25 for chat models, measured over 348 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 data1,291.1
252Qwen2.5 72B Instruct· LMArena data1,290.6
253Hunyuan TurboS20250226· LMArena data1,290.1
254Hunyuan Standard2025-02-10· LMArena data1,288.4
255Llama 3.1 Nemotron 70B Instruct· LMArena data1,287.5
256Llama 4 Scout· LMArena data1,286.1
257GPT 4o2024-05-13· LMArena data1,284.5
258Grok 22024-08-13· LMArena data1,283.9
259GPT 4o2024-08-06· LMArena data1,283.4
260GLM 4 Plus· LMArena data1,282.8
261Qwen Max0919· LMArena data1,279.6
262Gemini 1.5 Flash002· LMArena data1,279.5
263Granite 4.0 H Small· LMArena data1,278.6
264DeepSeek V2.51210· LMArena data1,278.4
265Gemini 1.5 Pro001· LMArena data1,276.4
266Llama 3.1 405B InstructFP8· LMArena data1,276.2
267Gemini Advanced0514· LMArena data1,276.1
268Llama 3.1 405B InstructBF16· LMArena data1,275.8
269GPT-4.1 nano2025-04-14· LMArena data1,274.7
270Gemma 3n E4B IT· LMArena data1,273.8
271DeepSeek V2.5· LMArena data1,273.7
272Grok 2 mini2024-08-13· LMArena data1,272.8
273GPT-4o mini2024-07-18· LMArena data1,271.6
274Mistral Small 3.1 24B Instruct2503· LMArena data1,271.2
275GPT-4 Turbo Preview1106· LMArena data1,270.3
276GPT-4 Turbo2024-04-09· LMArena data1,268.1
277Llama 3.3 70B Instruct· LMArena data1,267.3
278Mistral Large 22407· LMArena data1,265.6
279Claude 3 Opus20240229· LMArena data1,264
280Nova Pro 1.0· LMArena data1,261.4
281GPT-4 Turbo Preview0125· LMArena data1,261.2
282Mistral Large2411· LMArena data1,257.4
283Magistral Medium2506· LMArena data1,256.7
284Gemma 3 4B IT· LMArena data1,256.2
285Phi 4· LMArena data1,254.2
286Llama 3.1 70B Instruct· LMArena data1,253.6
287Claude Haiku 3.520241022· LMArena data1,249
288Llama 3.1 Tulu 3 70B· LMArena data1,246.4
289Hunyuan Standard 256K· LMArena data1,243.4
290Qwen 2.5 Coder 32B Instruct· LMArena data1,243.4
291Mistral Small 24B Instruct2501· LMArena data1,243.4
292Reka Core20240904· LMArena data1,239.2
293Gemini 1.5 Flash001· LMArena data1,238.2
294Nova Lite 1.0· LMArena data1,234.7
295Athene 70B0725· LMArena data1,233.4
296DeepSeek Coder V2 Instruct· LMArena data1,232.9
297Qwen2 72B Instruct· LMArena data1,230.8
298GLM 40520· LMArena data1,229.9
299Llama 3.1 Nemotron 51B Instruct· LMArena data1,228.6
300GPT 40314· LMArena data1,227.8

Showing 251–300 of 402 results.

Data credits: LMArena, CC BY 4.0 (licence)