Q1 Which AI gives the answers people prefer most? Arena Text · Top 14%#15 Q1 Which AI gives the answers people prefer most? Arena Text · Top 14%#15 Q2 On the old v1, could it solve problems it had not seen in training? LiveBench Overall v1 · Top 25%#13 Q3 How high was its intelligence index on the old v4 edition? AA Intelligence Index v4 · Top 32%#24
Q1 Top 14%#15 of 108 1490 Q2 78.8 Q3 39.6 About Arena Text More about Arena Text
Q1 Can it handle customer service by the rules? τ³-Banking · Top 26%#20 Q1 Can it handle customer service by the rules? τ³-Banking · Top 26%#20 Q2 Would a client accept its work on freelance jobs? Remote Labor Index · Bottom 34%#11
Q1 Top 26%#20 of 79 35.5% Q2 5.0% By effort Low 29.5% Medium 35.5% High 32.8%
About τ³-Banking More about τ³-Banking
Q1 Can it pull scattered facts from several long documents into one answer? Long Context Reasoning · Top 14%#16 Q1 Can it pull scattered facts from several long documents into one answer? Long Context Reasoning · Top 14%#16 Q2 Which AI gives the business answers people prefer most? Arena Business, Management & Finance · Top 15%#16 Q3 Can it do professional work in banking, consulting and law? APEX-Agents · Top 23%#13 Q1 Top 14%#16 of 121 83.0% Q2 1486 Q3 67.8% Q4 59.0% Q5 22.8% Q6 34.6% Q7 53.4% Q8 1476 Q9 68.0 Q10 23.8% Q11 8.8% By effort Low 78.7% Medium 83.0% High 81.7%
About Long Context Reasoning More about Long Context Reasoning
Q1 Which AI gives answers people prefer most, without false claims? Arena Text Factuality · Top 22%#22 Q1 Which AI gives answers people prefer most, without false claims? Arena Text Factuality · Top 22%#22 Q2 Does it know facts accurately without searching? SimpleQA Verified · Top 22%#15 Q3 Can it carry out scientists' real research tasks in a terminal? Terminal-Bench-Science 0.1 · Bottom 48%#22
Q1 Top 22%#22 of 103 1476 Q2 69.2% Q3 7.1% About Arena Text Factuality More about Arena Text Factuality
Q1 Can AI carry out long development tasks autonomously by operating a computer itself? Terminal-Bench 4.0 · Top 41%#34 Q1 Can AI carry out long development tasks autonomously by operating a computer itself? Terminal-Bench 4.0 · Top 41%#34 Q2 Which AI do people prefer most for coding answers? Arena Coding · Top 31%#33 Q3 Can AI test and fix code itself until it completes the given task? LiveBench Agentic Coding v1 · Top 24%#12 Q1 Top 41%#34 of 84 13.6% Q2 1521 Q3 58.3 Q4 904 Q5 85.8% Q6 76.1 Q7 1589 Q8 59.8% Q9 43.6% Q10 65.5% Q11 1513 Q12 78.9 Q13 0.0% Q14 70.4% Q15 20.3% Q16 41.1% About Terminal-Bench 4.0 More about Terminal-Bench 4.0
Q1 Which AI gives the writing and literature answers people prefer most? Arena Writing, Literature & Language · Top 10%#10 Q1 Which AI gives the writing and literature answers people prefer most? Arena Writing, Literature & Language · Top 10%#10 Q2 Which AI do people prefer most in long conversations? Arena Multi-turn · Top 18%#19 Q3 Which AI writes the creative pieces people prefer most? Arena Creative Writing · Top 10%#10 Q1 Top 10%#10 of 108 1491 Q2 1490 Q3 1487 Q4 1487 Q5 85.5 Q6 83.9% Q7 1475 Q8 79.9 About Arena Writing, Literature & Language More about Arena Writing, Literature & Language
Q1 Can it solve problems that stump experts across fields? Humanity's Last Exam · Top 14%#16 Q1 Can it solve problems that stump experts across fields? Humanity's Last Exam · Top 14%#16 Q2 Can it solve research-level physics problems? CritPt · Top 48%#41 Q3 Which AI gives the math answers people prefer most? Arena Math · Top 16%#16 Q1 Top 14%#16 of 122 47.9% Q2 14.3% Q3 1501 Q4 94.5% Q5 97.2% Q6 93.5 Q7 71.6% Q8 36.6% Q9 58.0% Q10 1513 Q11 1497 Q12 87.8 Q13 95.5% Q14 84.6% Q15 66.0% Q16 37.0% Q17 96.8% Q18 50.4 By effort Low 35.1% Medium 39.0% High 47.9%
About Humanity's Last Exam More about Humanity's Last Exam
Q1 Which AI gives the photo and chart answers people prefer most? Arena Vision · Top 10%#6 Q1 Which AI gives the photo and chart answers people prefer most? Arena Vision · Top 10%#6 Q2 Can it spot the wrong step from assembly photos? Furniture Assembly · Bottom 16%#33 Q3 Can it read charts and answer accurately? CharXiv · Top 26%#8
Q1 Top 10%#6 of 66 1314 Q2 26.7% Q3 88.7% About Arena Vision More about Arena Vision
Q1 Which AI do people prefer most when asking in Korean? Arena Korean · Top 8%#7 Q1 Which AI do people prefer most when asking in Korean? Arena Korean · Top 8%#7 Q2 Which AI do people prefer most when asking in Japanese? Arena Japanese · Top 5%#4 Q3 Which AI do people prefer most on tough questions? Arena Hard Prompts · Top 18%#19 Q1 Top 8%#7 of 91 1477 Q2 1520 Q3 1511 Q4 1519 Q5 1507 Q6 1497 Q7 75.5% Q8 49.8% Q9 78.7% Q10 80.9% Q11 65.4% Q12 94.9% Q13 94.5% Q14 94.8% Q15 1397 Q16 552 Q17 371 Q18 83.0% Q19 73.2% About Arena Korean More about Arena Korean
Q1 Does it call out a nonsense question instead of playing along? BullshitBench v2 · Bottom 43%#83 Q1 Does it call out a nonsense question instead of playing along? BullshitBench v2 · Bottom 43%#83 Q2 Can it sort word-puzzle tiles into groups using meaning and wordplay? NYT Connections Extended · Top 13%#14 Q3 What would it score on Korea's 2026 CSAT (Suneung)? CSAT 2026 · Top 14%#8 Q1 Bottom 43%#83 of 143 35.0% Reasoning effort: Minimal Q2 94.0% Q3 443 Q4 1006 Q5 1075 Q6 817 By effort Minimal 35.0% Max 31.0%
About BullshitBench v2 More about BullshitBench v2