Benchmark By Score Rank
Overall 1
AA Intelligence Index4.3.2 AA Intelligence Index 4.3.2: Artificial Analysis combines tests of knowledge, reasoning, coding and tool use into one score AA 40.9 #28 of 337
Knowledge 5
AA OmniscienceAccuracy AA Omniscience · Accuracy: Measures correct answers on the AA Omniscience factual knowledge test AA 54.6% #11 of 321+1 versions & slices Hallucination Rate AA Omniscience · Hallucination Rate: Measures how often attempted answers are incorrect on the AA Omniscience factual knowledge test AA 55.2% #99 of 321
AA Omniscience Index AA Omniscience Index: Scores factual knowledge while penalizing incorrect answers to encourage admitting uncertainty AA 29.6 #13 of 321
SimpleQA Verifiedv1 SimpleQA Verified v1: Accuracy on short factual questions with verified answers, answered without web access 69.7% #8 of 83
Mystery Game Puzzles Mystery Game Puzzles: Models choose the best move in puzzle positions from an undisclosed game Epoch 47% #13 of 64
Reasoning 6
ARC-AGI-1public evaluation ARC-AGI-1 · public-eval: Models infer transformations from examples and solve unseen grid puzzles; the evaluation split is explicit ARC 98% #10 of 61+1 versions & slices semi-private ARC-AGI-1 · semi-private: Models infer transformations from examples and solve unseen grid puzzles; the evaluation split is explicit ARC 98.5% #1 of 64
ARC-AGI-2public evaluation ARC-AGI-2 · public-eval: Models infer transformations from examples and solve unseen grid puzzles; the evaluation split is explicit ARC 92.5% #9 of 63+1 versions & slices semi-private ARC-AGI-2 · semi-private: Models infer transformations from examples and solve unseen grid puzzles; the evaluation split is explicit ARC 89.2% #9 of 66
ARC-AGI-3semi-private ARC-AGI-3 · semi-private: Measures interactive game solving through relative human action efficiency ARC 10.4% #5 of 18
Chess Puzzles Chess Puzzles: Models choose the best next move from a novel chess position Epoch 58% #10 of 68
Math 15
AA Analyst Agent AA Analyst Agent: Data-analyst tasks: the model explores data with tools and reports findings AA 46.3% #13 of 43
FrontierMath Erdos FrontierMath Erdos: Models solve expert-written advanced mathematics problems on the named tier and edition Epoch 0% #6 of 13
FrontierMath Tier 4 (v2) FrontierMath Tier 4 (v2): Models solve expert-written advanced mathematics problems on the named tier and edition Epoch 22% #38 of 54
FrontierMath Tiers 1–3 (v2) FrontierMath Tiers 1–3 (v2): Models solve expert-written advanced mathematics problems on the named tier and edition Epoch 68.4% #26 of 69
MathArenaArXivMath · April 2026 MathArena · arxiv · 04/2026: Answer accuracy on the named mathematics competition under the recorded effort MathArena 64.2% #6 of 17+10 versions & slices ArXivMath · August 2026 MathArena · arxiv · 08/2026: Answer accuracy on the named mathematics competition under the recorded effort MathArena 40.4% #9 of 12ArXivMath · June 2026 MathArena · arxiv · 06/2026: Answer accuracy on the named mathematics competition under the recorded effort MathArena 67.4% #15 of 24ArXivMath · May 2026 MathArena · arxiv · 05/2026: Answer accuracy on the named mathematics competition under the recorded effort MathArena 58.3% #12 of 19ArXivMath · September 2026 MathArena · arxiv · 09/2026: Answer accuracy on the named mathematics competition under the recorded effort MathArena 50% #4 of 7BrokenArXiv · April 2026 MathArena · arxiv_false · 04/2026: Answer accuracy on the named mathematics competition under the recorded effort MathArena 33.6% #7 of 15BrokenArXiv · August 2026 MathArena · arxiv_false · 08/2026: Answer accuracy on the named mathematics competition under the recorded effort MathArena 27.4% #10 of 12BrokenArXiv · June 2026 MathArena · arxiv_false · 06/2026: Answer accuracy on the named mathematics competition under the recorded effort MathArena 20.8% #18 of 24BrokenArXiv · May 2026 MathArena · arxiv_false · 05/2026: Answer accuracy on the named mathematics competition under the recorded effort MathArena 19% #13 of 18BrokenArXiv · September 2026 MathArena · arxiv_false · 09/2026: Answer accuracy on the named mathematics competition under the recorded effort MathArena 26.7% #7 of 7ArXivLean · June 2026 MathArena · arxivlean · 06/2026: Answer accuracy on the named mathematics competition under the recorded effort MathArena 19.6% #5 of 9
Science 5
Humanity's Last Exam Humanity's Last Exam: Very hard expert questions across many fields, written to stump today’s models AA 47.8% #17 of 337
CritPt (physics) CritPt (physics): Research-level physics problems that need many careful reasoning steps AA 18.3% #33 of 324
Vals Science Index Vals Science Index: Scientific problem solving across the index’s component evaluations Vals 35.8% #11 of 18
SciCode SciCode: Writing code for real scientific computing problems (physics, biology, math) AA 56.6% #20 of 140
Terminal-Bench Science0.1 Terminal-Bench Science 0.1: Agents complete expert-authored research workflows across five scientific domains in a terminal AA 10% #14 of 33
Coding 9
Terminal-Bench4.0 Terminal-Bench 4.0 · source setting 2: Agents use a terminal to complete executable coding, system administration and research tasks in the 4.0 suite 19.7% #30 of 132+1 versions & slices 2.1 Terminal-Bench 2.1 · source setting 2: Older, easier version of Terminal-Bench; Still widely reported AA 87.6% #8 of 156
DeepSWE1.1 DeepSWE 1.1: Agents change repository code to resolve software engineering tasks under the stated DeepSWE edition 73.8% #2 of 44
FrontierCode1.1 FrontierCode 1.1: Agents implement changes in real repositories, graded against maintainer-defined requirements AA 36.4% #21 of 28
ITBench SRE ITBench SRE: IT incident tasks: diagnose and fix broken systems like a site-reliability engineer AA 52.5% #3 of 37
WebDev Arena WebDev Arena: Human pairwise preferences in the stated webdev vote pool Arena 1,585 #28 of 97+1 versions & slices full cohort WebDev Arena · snapshot 2026-10-19 · native board settings: Human pairwise preferences in the stated webdev vote pool Arena 1,585 #28 of 97
SWE-bench Pro (v2 full) SWE-bench Pro (v2 full): Agents resolve repository issues under the network-locked v2 protocol; full and 51-task Hard results are separate Scale 94.9% #12 of 16
SWE-bench Pro (v2 hard) SWE-bench Pro (v2 hard): Agents resolve repository issues under the network-locked v2 protocol; full and 51-task Hard results are separate Scale 58.8% #13 of 16
Agents 5
AutomationBench-AA AutomationBench-AA: Agents orchestrate SaaS workflows through APIs; objectives receive credit only when the task has no guardrail violation AA 59.9% #18 of 133
AA-Briefcase1.1 · Elo AA-Briefcase 1.1 · Elo: Agents produce professional deliverables; blind pairwise comparisons produce the source Elo rating AA 1,203 #43 of 140
GDPval-AA (work tasks)2.1 GDPval-AA (work tasks) 2.1: Real work tasks from many professions (reports, spreadsheets, analyses), graded against expert work AA 46.8 #35 of 175
Furniture Assembly Furniture Assembly: Models inspect furniture assembly photos, identify incorrect builds and locate the mistaken step Epoch 31.7% #23 of 31
τ³-BankingAA τ³-Banking · AA: Agents retrieve banking knowledge and execute tools to resolve customer-support workflows AA 44.9% #11 of 134
Multimodal 3
MMMU-Pro (vision) MMMU-Pro (vision): College-level questions that require understanding images: charts, diagrams, photos AA 85.6% #4 of 127
Arena Vision Arena Vision · overall: Human pairwise preferences in the stated vision vote pool Arena 1,315 #7 of 88
Chartography Chartography: Expert chart reasoning over professional scientific, medical, financial and engineering charts Surge 40.9% #7 of 37
Writing 18
Arena Textcreative writing Arena Text · creative writing: Human pairwise preferences in the stated text vote pool Arena 1,487 #6 of 195+14 versions & slices overall Arena Text · overall: Human pairwise preferences in the stated text vote pool Arena 1,496 #8 of 196Chinese Arena Text · chinese: Human pairwise preferences in the stated text vote pool Arena 1,551 #5 of 181English Arena Text · english: Human pairwise preferences in the stated text vote pool Arena 1,499 #8 of 188French Arena Text · french: Human pairwise preferences in the stated text vote pool Arena 1,492 #16 of 149German Arena Text · german: Human pairwise preferences in the stated text vote pool Arena 1,498 #6 of 154instruction following Arena Text · instruction following: Human pairwise preferences in the stated text vote pool Arena 1,491 #8 of 188Japanese Arena Text · japanese: Human pairwise preferences in the stated text vote pool Arena 1,506 #4 of 138Korean Arena Text · korean: Human pairwise preferences in the stated text vote pool Arena 1,462 #9 of 144non-English Arena Text · non english: Human pairwise preferences in the stated text vote pool Arena 1,491 #5 of 188Polish Arena Text · polish: Human pairwise preferences in the stated text vote pool Arena 1,498 #3 of 133Russian Arena Text · russian: Human pairwise preferences in the stated text vote pool Arena 1,514 #5 of 186Spanish Arena Text · spanish: Human pairwise preferences in the stated text vote pool Arena 1,486 #10 of 155overall · full cohort Arena Text · overall · snapshot 2026-10-15 · native board settings: Human pairwise preferences in the stated text vote pool Arena 1,499 #7 of 188creative writing · full cohort Arena Text · creative writing · snapshot 2026-10-15 · native board settings: Human pairwise preferences in the stated text vote pool Arena 1,492 #7 of 187
EQ-Bench Longform Creative Writing1.11 Longform Creative Writing 1.11: Models plan, revise and write a story across eight turns, graded with a writing rubric EQ-Bench 76.8% #24 of 95
EQ-Bench Creative Writingv3 Creative Writing v3 · Elo: Judge-rated creative writing under the v3 prompt set and pairwise ranking EQ-Bench 1,748 #28 of 91
Lech Mazur Short-Story Lech Mazur Short-Story · comparison score: Judges compare stories written to ten required elements; the comparison score can be negative Lech 0.1 #22 of 44
Health & law 6
Harvey LAB-AA1.1 Harvey LAB-AA 1.1: Legal agents work from case documents; every rubric criterion must pass and material hallucinations receive no credit AA 0% #18 of 23
MLCR-AA MLCR-AA: Models synthesize long medical records; the headline credits concise answers judged both accurate and complete AA 21.7% #20 of 64
Vals LegalBench Vals LegalBench: Legal reasoning and legal-language understanding on LegalBench Vals 87% #5 of 89
Vals MedCode Vals MedCode: Medical coding accuracy on patient encounters Vals 48.1% #26 of 78
Vals MedScribe Vals MedScribe: Accuracy of clinical notes produced from consultation transcripts Vals 84.5% #30 of 80
Vals Legal Research Bench Vals Legal Research Bench: Grounded legal research; a task passes only if all criteria pass Vals 38.9% #27 of 69
Documents 2
AA-LCR (long context)1.1 AA-LCR (long context) 1.1: Questions that need reading and connecting facts across very long documents (around 100K tokens) AA 81.3% #32 of 275
GDP.pdfAA all-pass GDP.pdf · AA all-pass: Models reason over professional PDF documents; an attempt passes only when every atomic criterion passes 21% #22 of 133
Legacy · saturated or no longer reported 3
GPQA Diamond GPQA Diamond: Graduate-level biology, chemistry and physics questions that can’t be solved by a quick web search 95.3% #2 of 315+1 versions & slices Epoch GPQA Diamond · source setting 2: Models answer graduate-level science multiple-choice questions Epoch 95.4% #4 of 133
OTIS Mock AIME 20242025 OTIS Mock AIME 2024 2025: Models solve the stated OTIS mock AIME mathematics problems Epoch 98.9% #15 of 130
No benchmark matches
CheaperStep 5 PreviewthinkingOpen+6%-17%$1.03No PDF input+4
CheaperGLM 5.3 FlashmaxOpen-2%-77%$0.28Defaults to max · Self-hosted+4
CheaperGPT-6 Solmax-6%-16%$1.04
OpenKimi K3max-9%+61%$2.00Sampling parameters are fixed+3