Skip to content

Benchmark

Arena ELO - Math

Math benchmarks measure how well models solve maths problems. Higher scores are better.

Models measured
350
Scores recorded
405
Sources
1
IQ weight
1.24

Score distribution

1,316
Mean
1,367.4
Median
152.6
Std dev
1,089
10th percentile
1,474.8
90th percentile
837.8
Min
837.81,547.6
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.24 for chat models, measured over 350 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
201Qwen2.5-Max· LMArena data1,369.1
202Kimi K2 Preview0711· LMArena data1,367.4
203Step3· LMArena data1,365.3
204Qwen3 Coder· LMArena data1,365.1
205Trinity Large Thinking· LMArena data1,365.1
206Nemotron 3.5 Lightning 30B A3BNVFP4· LMArena data1,365
207Ling Flash 2.0· LMArena data1,363.8
208Gemini 2.5 Flash-Lite PreviewThinking, 06-17· LMArena data1,362.3
209MiniMax M1· LMArena data1,361
210Llama 3.1 Nemotron Ultra 253B v1· LMArena data1,359.5
211QwQ 32B· LMArena data1,358.9
212Claude Sonnet 420250514· LMArena data1,358.9
213Hunyuan TurboS20250416· LMArena data1,358.5
214Nova 2 Lite· LMArena data1,358.2
215O1 Mini· LMArena data1,358.2
216GLM 4.7 Flash· LMArena data1,355.3
217Qwen3 30B A3B· LMArena data1,354.7
218GLM 4.5V· LMArena data1,353.7
219Mistral Medium 32505· LMArena data1,352.3
220MiniMax M2· LMArena data1,352.3
221Gemini 2.0 Flash001· LMArena data1,351.9
222Ring Flash 2.0· LMArena data1,347.7
223GPT-4.1 mini2025-04-14· LMArena data1,342.9
224Mistral Small 3.22506· LMArena data1,341.9
225Claude Sonnet 3.7Thinking 32K, 20250219· LMArena data1,336.5
226Trinity Large Preview· LMArena data1,335.8
227Qwen Plus0125· LMArena data1,325.9
228Claude Sonnet 3.720250219· LMArena data1,318.7
229GPT-5 nanoHigh· LMArena data1,318.3
230Step-1o Turbo202506· LMArena data1,317.9
231GPT OSS 20b· LMArena data1,317.1
232Olmo 3 32B Think· LMArena data1,315.5
233Gemini 1.5 Pro002· LMArena data1,314.8
234Granite 4.1 8B· LMArena data1,312.8
235Gemma 3 27B IT· LMArena data1,311.9
236DeepSeek V3· LMArena data1,310.7
237Gemini 2.0 Flash-Lite Preview02-05· LMArena data1,309.2
238Olmo 3.1 32B Instruct· LMArena data1,307.9
239Gemma 3 12B IT· LMArena data1,307.4
240Claude 3.5 Sonnet20241022· LMArena data1,306.6
241Olmo 3.1 32B Think· LMArena data1,304.6
242Step-2 16K Exp202412· LMArena data1,304.5
243Claude 3.5 Sonnet20240620· LMArena data1,303.4
244Athene V2 Chat· LMArena data1,300.3
245Command A03-2025· LMArena data1,300.1
246Yi Lightning· LMArena data1,299.7
247Llama 4 Maverick· LMArena data1,299.4
248Qwen2.5-Plus1127· LMArena data1,297.5
249Hunyuan TurboS20250226· LMArena data1,294
250DeepSeek V2.51210· LMArena data1,288.5

Showing 201–250 of 405 results.

Data credits: LMArena, CC BY 4.0 (licence)