Benchmark By Score Rank
Overall 1
AA Intelligence Index4.3.2 Artificial Analysis combines tests of knowledge, reasoning, coding and tool use into one score AA 34 #46 of 339
Knowledge 5
AA OmniscienceAccuracy Measures correct answers on the AA Omniscience factual knowledge test AA 50% #20 of 323+1 versions & slices Hallucination Rate Measures how often attempted answers are incorrect on the AA Omniscience factual knowledge test AA 55.6% #102 of 323
AA Omniscience Index Scores factual knowledge while penalizing incorrect answers to encourage admitting uncertainty AA 22.1 #24 of 323
SimpleQA Verifiedv1 Accuracy on short factual questions with verified answers, answered without web access 66.2% #11 of 84
Mystery Game Puzzles Models choose the best move in puzzle positions from an undisclosed game Epoch 30% #24 of 65
Reasoning 5
ARC-AGI-1public evaluation Models infer transformations from examples and solve unseen grid puzzles; the evaluation split is explicit ARC 92.8% #29 of 64+1 versions & slices semi-private Models infer transformations from examples and solve unseen grid puzzles; the evaluation split is explicit ARC 83.2% #38 of 67
ARC-AGI-2public evaluation Models infer transformations from examples and solve unseen grid puzzles; the evaluation split is explicit ARC 56.5% #32 of 66+1 versions & slices semi-private Models infer transformations from examples and solve unseen grid puzzles; the evaluation split is explicit ARC 50.4% #36 of 69
Chess Puzzles Models choose the best next move from a novel chess position Epoch 42% #27 of 69
Math 18
MathArenaAIME 2026 Answer accuracy on the named mathematics competition under the recorded effort MathArena 96.7% #6 of 24+15 versions & slices HMMT February 2026 Answer accuracy on the named mathematics competition under the recorded effort MathArena 89.4% #13 of 24Apex Answer accuracy on the named mathematics competition under the recorded effort MathArena 26% #11 of 34Apex Shortlist Answer accuracy on the named mathematics competition under the recorded effort MathArena 71.3% #13 of 30ArXivMath · April 2026 Answer accuracy on the named mathematics competition under the recorded effort MathArena 52.5% #11 of 17ArXivMath · June 2026 Answer accuracy on the named mathematics competition under the recorded effort MathArena 56.9% #19 of 24ArXivMath · May 2026 Answer accuracy on the named mathematics competition under the recorded effort MathArena 54.2% #15 of 19BrokenArXiv · April 2026 Answer accuracy on the named mathematics competition under the recorded effort MathArena 16% #13 of 15BrokenArXiv · June 2026 Answer accuracy on the named mathematics competition under the recorded effort MathArena 12% #23 of 24BrokenArXiv · May 2026 Answer accuracy on the named mathematics competition under the recorded effort MathArena 4% #18 of 18Kangaroo 2025 1-2 Answer accuracy on the named mathematics competition under the recorded effort MathArena 83.3% #6 of 17Kangaroo 2025 11-12 Answer accuracy on the named mathematics competition under the recorded effort MathArena 96.7% #6 of 17Kangaroo 2025 3-4 Answer accuracy on the named mathematics competition under the recorded effort MathArena 68.8% #6 of 17Kangaroo 2025 5-6 Answer accuracy on the named mathematics competition under the recorded effort MathArena 88.3% #2 of 17Kangaroo 2025 7-8 Answer accuracy on the named mathematics competition under the recorded effort MathArena 90% #6 of 16Kangaroo 2025 9-10 Answer accuracy on the named mathematics competition under the recorded effort MathArena 100% #1 of 16
FrontierMath Tier 4 (v2) Models solve expert-written advanced mathematics problems on the named tier and edition Epoch 22% #39 of 55
FrontierMath Tiers 1–3 (v2) Models solve expert-written advanced mathematics problems on the named tier and edition Epoch 58.9% #37 of 70
Science 3
Humanity's Last Exam Very hard expert questions across many fields, written to stump today’s models AA 40.8% #43 of 339
CritPt (physics) Research-level physics problems that need many careful reasoning steps AA 10.6% #58 of 326
SciCode Writing code for real scientific computing problems (physics, biology, math) AA 53.4% #39 of 142
Coding 6
Terminal-Bench4.0 Agents use a terminal to complete executable coding, system administration and research tasks in the 4.0 suite 7.1% #43 of 134+1 versions & slices 2.1 Older, easier version of Terminal-Bench; Still widely reported AA 77.5% #28 of 156
DeepSWE1.1 Agents change repository code to resolve software engineering tasks under the stated DeepSWE edition 46.7% #28 of 44
WebDev Arena Human pairwise preferences in the stated webdev vote pool Arena 1,538 #39 of 97+2 versions & slices October 23, 2026 Human pairwise preferences in the stated webdev vote pool Arena 1,538 #40 of 98October 19, 2026 Human pairwise preferences in the stated webdev vote pool Arena 1,538 #39 of 97
Agents 5
AutomationBench-AA Agents orchestrate SaaS workflows through APIs; objectives receive credit only when the task has no guardrail violation AA 53% #35 of 135
AA-Briefcase1.1 · Elo Agents produce professional deliverables; blind pairwise comparisons produce the source Elo rating AA 955 #62 of 142
GDPval-AA (work tasks)2.1 Real work tasks from many professions (reports, spreadsheets, analyses), graded against expert work AA 1,286 #47 of 177
Furniture Assembly Models inspect furniture assembly photos, identify incorrect builds and locate the mistaken step Epoch 23.3% #26 of 31
τ³-BankingAA Agents retrieve banking knowledge and execute tools to resolve customer-support workflows AA 29.9% #35 of 134
Multimodal 3
MMMU-Pro (vision) College-level questions that require understanding images: charts, diagrams, photos AA 83.2% #8 of 128
Arena Vision Human pairwise preferences in the stated vision vote pool Arena 1,299 #12 of 88
Chartography Expert chart reasoning over professional scientific, medical, financial and engineering charts Surge 34% #9 of 37
Writing 18
Arena Textcreative writing · October 22, 2026 Human pairwise preferences in the stated text vote pool Arena 1,466 #13 of 197+16 versions & slices creative writing Human pairwise preferences in the stated text vote pool Arena 1,471 #10 of 195overall Human pairwise preferences in the stated text vote pool Arena 1,483 #18 of 196Chinese Human pairwise preferences in the stated text vote pool Arena 1,531 #14 of 198English Human pairwise preferences in the stated text vote pool Arena 1,482 #15 of 205French Human pairwise preferences in the stated text vote pool Arena 1,504 #8 of 164German Human pairwise preferences in the stated text vote pool Arena 1,481 #14 of 167instruction following Human pairwise preferences in the stated text vote pool Arena 1,467 #23 of 205Japanese Human pairwise preferences in the stated text vote pool Arena 1,475 #12 of 152Korean Human pairwise preferences in the stated text vote pool Arena 1,430 #24 of 156non-English Human pairwise preferences in the stated text vote pool Arena 1,470 #16 of 205Polish Human pairwise preferences in the stated text vote pool Arena 1,475 #13 of 142Russian Human pairwise preferences in the stated text vote pool Arena 1,488 #12 of 203Spanish Human pairwise preferences in the stated text vote pool Arena 1,476 #14 of 168overall · October 15, 2026 Human pairwise preferences in the stated text vote pool Arena 1,479 #14 of 188creative writing · October 15, 2026 Human pairwise preferences in the stated text vote pool Arena 1,466 #12 of 187overall · October 22, 2026 Human pairwise preferences in the stated text vote pool Arena 1,480 #14 of 198
EQ-Bench Creative Writingv3 Creative Writing v3 · Elo: Judge-rated creative writing under the v3 prompt set and pairwise ranking EQ-Bench 1,602 #43 of 94
Health & law 5
MLCR-AA Models synthesize long medical records; the headline credits concise answers judged both accurate and complete AA 10% #57 of 103
Vals LegalBench Legal reasoning and legal-language understanding on LegalBench Vals 86.7% #9 of 89
Vals MedCode Medical coding accuracy on patient encounters Vals 53.2% #9 of 78
Vals MedScribe Accuracy of clinical notes produced from consultation transcripts Vals 79.7% #51 of 80
Vals Legal Research Bench Grounded legal research; a task passes only if all criteria pass Vals 25% #51 of 73
Documents 3
AA-LCR (long context)1.1 Questions that need reading and connecting facts across very long documents (around 100K tokens) AA 80% #44 of 317
GDP.pdfAA all-pass Models reason over professional PDF documents; an attempt passes only when every atomic criterion passes 17.4% #33 of 135
Arena Document Human pairwise preferences in the stated document vote pool Arena 1,456 #18 of 33
Legacy · saturated or no longer reported 3
GPQA Diamond Graduate-level biology, chemistry and physics questions that can’t be solved by a quick web search 92.8% #14 of 315+1 versions & slices Epoch Models answer graduate-level science multiple-choice questions Epoch 94.1% #9 of 134
OTIS Mock AIME 20242025 Models solve the stated OTIS mock AIME mathematics problems Epoch 94.2% #37 of 131
No benchmark matches
CheaperGPT-5.6 Lunamax-8%-70%$0.18
OpenKimi K3max-22%+300%$2.00Sampling parameters are fixed+3
Qwen3.8 FlashxhighOpen-49%-85%$0.15 / $0.47Context: 1M → 262K+6
Muse Glimmer 30BxhighOpen-49%-67%$0.30 / $1.10Context: 1M → 131K
GLM 5.3 FlashmaxOpen-52%-84%$0.28Defaults to max · Self-hosted+4