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 38.8 #35 of 337
Knowledge 4
AA OmniscienceAccuracy AA Omniscience · Accuracy: Measures correct answers on the AA Omniscience factual knowledge test AA 51.6% #17 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 54.1% #98 of 321
AA Omniscience Index AA Omniscience Index: Scores factual knowledge while penalizing incorrect answers to encourage admitting uncertainty AA 25.3 #20 of 321
SimpleQA Verifiedv1 SimpleQA Verified v1: Accuracy on short factual questions with verified answers, answered without web access 48.3% #29 of 83
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 94.8% #26 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 85.7% #33 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 58.2% #29 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 52.6% #33 of 66
ARC-AGI-3semi-private ARC-AGI-3 · semi-private: Measures interactive game solving through relative human action efficiency ARC 0.3% #15 of 18
Chess Puzzles Chess Puzzles: Models choose the best next move from a novel chess position Epoch 35% #38 of 68
Math 5
AA Analyst Agent AA Analyst Agent: Data-analyst tasks: the model explores data with tools and reports findings AA 35% #25 of 43
FrontierMath Tier 4 (v2) FrontierMath Tier 4 (v2): Models solve expert-written advanced mathematics problems on the named tier and edition Epoch 24.4% #37 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 57.2% #38 of 69
MathArenaArXivMath · June 2026 MathArena · arxiv · 06/2026: Answer accuracy on the named mathematics competition under the recorded effort MathArena 66.7% #16 of 24+1 versions & slices BrokenArXiv · June 2026 MathArena · arxiv_false · 06/2026: Answer accuracy on the named mathematics competition under the recorded effort MathArena 29.3% #16 of 24
Science 3
Humanity's Last Exam Humanity's Last Exam: Very hard expert questions across many fields, written to stump today’s models AA 42.7% #33 of 337
CritPt (physics) CritPt (physics): Research-level physics problems that need many careful reasoning steps AA 15.4% #43 of 324
SciCode SciCode: Writing code for real scientific computing problems (physics, biology, math) AA 55% #27 of 140
Coding 5
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 10.6% #39 of 132+1 versions & slices 2.1 Terminal-Bench 2.1 · source setting 2: Older, easier version of Terminal-Bench; Still widely reported AA 81.6% #19 of 156
DeepSWE1.1 DeepSWE 1.1: Agents change repository code to resolve software engineering tasks under the stated DeepSWE edition 53.8% #23 of 44
WebDev Arena WebDev Arena: Human pairwise preferences in the stated webdev vote pool Arena 1,553 #34 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 1,553 #34 of 97
Agents 7
AutomationBench-AA AutomationBench-AA: Agents orchestrate SaaS workflows through APIs; objectives receive credit only when the task has no guardrail violation AA 57.9% #24 of 133
AA-Briefcase1.1 · Elo AA-Briefcase 1.1 · Elo: Agents produce professional deliverables; blind pairwise comparisons produce the source Elo rating AA 1,281 #38 of 140
EnterpriseOps-Gym-AA EnterpriseOps-Gym-AA: Agents complete stateful enterprise workflows in eight domains; final database state determines success AA 40.8% #20 of 43
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 44.2 #40 of 175
τ³-benchbanking · 1.0.1 τ-bench official · 3-banking-1.0.1: Customer-service agents converse with simulated users, retrieve domain knowledge and change state through tools Sierra 47.9% #2 of 21
Furniture Assembly Furniture Assembly: Models inspect furniture assembly photos, identify incorrect builds and locate the mistaken step Epoch 22.5% #27 of 31
τ³-BankingAA τ³-Banking · AA: Agents retrieve banking knowledge and execute tools to resolve customer-support workflows AA 42.1% #15 of 134
Multimodal 3
MMMU-Pro (vision) MMMU-Pro (vision): College-level questions that require understanding images: charts, diagrams, photos AA 80.4% #19 of 127
Arena Vision Arena Vision · overall: Human pairwise preferences in the stated vision vote pool Arena 1,287 #20 of 88
Chartography Chartography: Expert chart reasoning over professional scientific, medical, financial and engineering charts Surge 17% #24 of 37
Writing 17
Arena Textcreative writing Arena Text · creative writing: Human pairwise preferences in the stated text vote pool Arena 1,452 #24 of 195+14 versions & slices overall Arena Text · overall: Human pairwise preferences in the stated text vote pool Arena 1,466 #34 of 196Chinese Arena Text · chinese: Human pairwise preferences in the stated text vote pool Arena 1,495 #44 of 181English Arena Text · english: Human pairwise preferences in the stated text vote pool Arena 1,453 #44 of 188French Arena Text · french: Human pairwise preferences in the stated text vote pool Arena 1,457 #46 of 149German Arena Text · german: Human pairwise preferences in the stated text vote pool Arena 1,448 #38 of 154instruction following Arena Text · instruction following: Human pairwise preferences in the stated text vote pool Arena 1,446 #38 of 188Japanese Arena Text · japanese: Human pairwise preferences in the stated text vote pool Arena 1,431 #27 of 138Korean Arena Text · korean: Human pairwise preferences in the stated text vote pool Arena 1,406 #38 of 144non-English Arena Text · non english: Human pairwise preferences in the stated text vote pool Arena 1,440 #36 of 188Polish Arena Text · polish: Human pairwise preferences in the stated text vote pool Arena 1,433 #45 of 133Russian Arena Text · russian: Human pairwise preferences in the stated text vote pool Arena 1,449 #37 of 186Spanish Arena Text · spanish: Human pairwise preferences in the stated text vote pool Arena 1,451 #41 of 155overall · full cohort Arena Text · overall · snapshot 2026-10-15 · native board settings: Human pairwise preferences in the stated text vote pool Arena 1,448 #38 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 1,442 #28 of 187
EQ-Bench Creative Writingv3 Creative Writing v3 · Elo: Judge-rated creative writing under the v3 prompt set and pairwise ranking EQ-Bench 1,579 #44 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 -5.1 #44 of 44
Health & law 5
MLCR-AA MLCR-AA: Models synthesize long medical records; the headline credits concise answers judged both accurate and complete AA 15% #34 of 64
Vals LegalBench Vals LegalBench: Legal reasoning and legal-language understanding on LegalBench Vals 86% #15 of 89
Vals MedCode Vals MedCode: Medical coding accuracy on patient encounters Vals 43.3% #38 of 78
Vals MedScribe Vals MedScribe: Accuracy of clinical notes produced from consultation transcripts Vals 86.9% #18 of 80
Vals Legal Research Bench Vals Legal Research Bench: Grounded legal research; a task passes only if all criteria pass Vals 38% #32 of 69
Documents 3
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 79.3% #53 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 18.8% #27 of 133
Arena Document Arena Document · overall: Human pairwise preferences in the stated document vote pool Arena 1,452 #20 of 33
Legacy · saturated or no longer reported 9
GPQA Diamond GPQA Diamond: Graduate-level biology, chemistry and physics questions that can’t be solved by a quick web search 93.1% #12 of 315+1 versions & slices Epoch GPQA Diamond · source setting 2: Models answer graduate-level science multiple-choice questions Epoch 93.4% #13 of 133
OTIS Mock AIME 20242025 OTIS Mock AIME 2024 2025: Models solve the stated OTIS mock AIME mathematics problems Epoch 97.8% #26 of 130
SWE-rebenchall · 2026-05-15 to 2026-07-01 SWE-rebench · all · 2026-05-15 to 2026-07-01: Repository issue resolution over the exact fresh pull-request window, under the source agent version Nebius 63.8% #2 of 6+5 versions & slices Go · 2026-05-15 to 2026-07-01 SWE-rebench · go · 2026-05-15 to 2026-07-01: Repository issue resolution over the exact fresh pull-request window, under the source agent version Nebius 68.6% #4 of 6Java · 2026-05-15 to 2026-07-01 SWE-rebench · java · 2026-05-15 to 2026-07-01: Repository issue resolution over the exact fresh pull-request window, under the source agent version Nebius 60% #1 of 6Python · 2026-05-15 to 2026-07-01 SWE-rebench · python · 2026-05-15 to 2026-07-01: Repository issue resolution over the exact fresh pull-request window, under the source agent version Nebius 54% #2 of 6Rust · 2026-05-15 to 2026-07-01 SWE-rebench · rust · 2026-05-15 to 2026-07-01: Repository issue resolution over the exact fresh pull-request window, under the source agent version Nebius 64% #2 of 6TypeScript · 2026-05-15 to 2026-07-01 SWE-rebench · typescript · 2026-05-15 to 2026-07-01: Repository issue resolution over the exact fresh pull-request window, under the source agent version Nebius 70.8% #3 of 6
No benchmark matches
CheaperGPT-6.1 Solmax+64%-18%$0.72Remove unsupported parameters+5
OpenMuse Spark Openxhighsame-4%$0.84Reasoning off leaves output unchanged · Self-hosted / llama.cpp+1
CheaperClaude Opus 5.5low+54%-37%$0.55No forced tool use+5
Gemini 3.8 Flashmedium+13%+6%*$0.93Sampling parameters are ignored+1
Kimi K3maxOpen-22%+127%$2.00Sampling parameters are fixed+3
Step 5 PreviewthinkingOpen-35%+17%$1.03web_search not available · Chat Completions+2
GLM 5.3maxOpen-44%+128%$2.01No image input+4MMotif 3Open-60%-71%$0.50 / $2Context: 500K → 262K+1