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 25.7 #79 of 337
Knowledge 5
AA OmniscienceAccuracy Measures correct answers on the AA Omniscience factual knowledge test AA 33.2% #64 of 321+1 versions & slices Hallucination Rate Measures how often attempted answers are incorrect on the AA Omniscience factual knowledge test AA 63% #115 of 321
AA Omniscience Index Scores factual knowledge while penalizing incorrect answers to encourage admitting uncertainty AA -8.9 #90 of 321
SimpleQA Verifiedv1 Accuracy on short factual questions with verified answers, answered without web access 19.1% #70 of 83
Mystery Game Puzzles Models choose the best move in puzzle positions from an undisclosed game Epoch 6% #63 of 64
Reasoning 4
ARC-AGI-1public evaluation Models infer transformations from examples and solve unseen grid puzzles; the evaluation split is explicit ARC 89.4% #32 of 61+1 versions & slices semi-private Models infer transformations from examples and solve unseen grid puzzles; the evaluation split is explicit ARC 84% #35 of 64
ARC-AGI-2public evaluation Models infer transformations from examples and solve unseen grid puzzles; the evaluation split is explicit ARC 39.3% #31 of 63+1 versions & slices semi-private Models infer transformations from examples and solve unseen grid puzzles; the evaluation split is explicit ARC 40.1% #35 of 66
Math 3
AA Analyst Agent Data-analyst tasks: the model explores data with tools and reports findings AA 27.5% #27 of 43
FrontierMath Tier 4 (v2) Models solve expert-written advanced mathematics problems on the named tier and edition Epoch 17.1% #42 of 54
FrontierMath Tiers 1–3 (v2) Models solve expert-written advanced mathematics problems on the named tier and edition Epoch 46.3% #48 of 69
Science 4
Humanity's Last Exam Very hard expert questions across many fields, written to stump today’s models AA 33.3% #72 of 337
CritPt (physics) Research-level physics problems that need many careful reasoning steps AA 8.3% #68 of 324
SciCode Writing code for real scientific computing problems (physics, biology, math) AA 49.7% #52 of 140
Terminal-Bench Science0.1 Agents complete expert-authored research workflows across five scientific domains in a terminal AA 0% #27 of 33
Coding 4
Terminal-Bench4.0 Agents use a terminal to complete executable coding, system administration and research tasks in the 4.0 suite 1% #54 of 132+1 versions & slices 2.1 Older, easier version of Terminal-Bench; Still widely reported AA 55.1% #62 of 156
WebDev Arena Human pairwise preferences in the stated webdev vote pool Arena 1,409 #65 of 97+1 versions & slices full cohort Human pairwise preferences in the stated webdev vote pool Arena 1,409 #65 of 97
Agents 5
AutomationBench-AA Agents orchestrate SaaS workflows through APIs; objectives receive credit only when the task has no guardrail violation AA 4.8% #86 of 133
AA-Briefcase1.1 · Elo Agents produce professional deliverables; blind pairwise comparisons produce the source Elo rating AA 912 #66 of 140
EnterpriseOps-Gym-AA Agents complete stateful enterprise workflows in eight domains; final database state determines success AA 42.7% #17 of 43
GDPval-AA (work tasks)2.1 Real work tasks from many professions (reports, spreadsheets, analyses), graded against expert work AA 31.2 #59 of 175
τ³-BankingAA Agents retrieve banking knowledge and execute tools to resolve customer-support workflows AA 18.8% #59 of 134
Multimodal 3
MMMU-Pro (vision) College-level questions that require understanding images: charts, diagrams, photos AA 74% #57 of 127
Arena Vision Human pairwise preferences in the stated vision vote pool Arena 1,237 #52 of 88
Chartography Expert chart reasoning over professional scientific, medical, financial and engineering charts Surge 9.1% #36 of 37
Writing 16
Arena Textcreative writing Human pairwise preferences in the stated text vote pool Arena 1,314 #120 of 195+14 versions & slices overall Human pairwise preferences in the stated text vote pool Arena 1,404 #94 of 196Chinese Human pairwise preferences in the stated text vote pool Arena 1,463 #69 of 181English Human pairwise preferences in the stated text vote pool Arena 1,421 #82 of 188French Human pairwise preferences in the stated text vote pool Arena 1,433 #73 of 149German Human pairwise preferences in the stated text vote pool Arena 1,407 #70 of 154instruction following Human pairwise preferences in the stated text vote pool Arena 1,399 #81 of 188Japanese Human pairwise preferences in the stated text vote pool Arena 1,401 #44 of 138Korean Human pairwise preferences in the stated text vote pool Arena 1,364 #67 of 144non-English Human pairwise preferences in the stated text vote pool Arena 1,400 #82 of 188Polish Human pairwise preferences in the stated text vote pool Arena 1,426 #53 of 133Russian Human pairwise preferences in the stated text vote pool Arena 1,391 #89 of 186Spanish Human pairwise preferences in the stated text vote pool Arena 1,422 #69 of 155overall · full cohort Human pairwise preferences in the stated text vote pool Arena 1,413 #81 of 188creative writing · full cohort Human pairwise preferences in the stated text vote pool Arena 1,331 #105 of 187
EQ-Bench Creative Writingv3 Creative Writing v3 · Elo: Judge-rated creative writing under the v3 prompt set and pairwise ranking EQ-Bench 1,491 #50 of 91
Health & law 5
Harvey LAB-AA1.1 Legal agents work from case documents; every rubric criterion must pass and material hallucinations receive no credit AA 0% #18 of 23
Vals LegalBench Legal reasoning and legal-language understanding on LegalBench Vals 83% #48 of 89
Vals MedCode Medical coding accuracy on patient encounters Vals 37.9% #57 of 78
Vals MedScribe Accuracy of clinical notes produced from consultation transcripts Vals 84.1% #33 of 80
Vals Legal Research Bench Grounded legal research; a task passes only if all criteria pass Vals 25.5% #45 of 69
Documents 2
AA-LCR (long context)1.1 Questions that need reading and connecting facts across very long documents (around 100K tokens) AA 75.7% #86 of 275
GDP.pdfAA all-pass Models reason over professional PDF documents; an attempt passes only when every atomic criterion passes 11.2% #52 of 133
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 89.5% #41 of 315+1 versions & slices Epoch Models answer graduate-level science multiple-choice questions Epoch 88.5% #40 of 133
OTIS Mock AIME 20242025 Models solve the stated OTIS mock AIME mathematics problems Epoch 90% #47 of 130
No benchmark matches
CheaperGPT-5.6 Lunamax+77%-29%$0.18
OpenGLM 5.3 Flashmax+20%-63%$0.28Memory (4-bit): 160 GB → 193 GB+5
CheaperClaude Haiku 5.5max+62%-69%$0.21Remove sampling parameters+5
DeepSeek V4 Pro 0813maxOpen+39%+211%$0.67Memory (4-bit): 160 GB → 990 GB+6
Qwen3.6 27BunstatedOpen-24%+112%$0.60 / $3.60