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 50.9 #9 of 373
Knowledge 5 Leads on 1
AA OmniscienceAccuracy Measures correct answers on the AA Omniscience factual knowledge test AA 61.1% #5 of 339+1 versions & slices Hallucination Rate Measures how often attempted answers are incorrect on the AA Omniscience factual knowledge test AA 44.8% #77 of 339
AA Omniscience Index Scores factual knowledge while penalizing incorrect answers to encourage admitting uncertainty AA 43.7 #2 of 339
SimpleQA Verifiedv1 Accuracy on short factual questions with verified answers, answered without web access 75.8% #3 of 84
Mystery Game Puzzles Models choose the best move in puzzle positions from an undisclosed game Epoch 84% #1 of 65
Reasoning 6 Leads on 2
ARC-AGI-1public evaluation Models infer transformations from examples and solve unseen grid puzzles; the evaluation split is explicit ARC 98.5% #6 of 64+1 versions & slices semi-private Models infer transformations from examples and solve unseen grid puzzles; the evaluation split is explicit ARC 98.5% #1 of 67
ARC-AGI-2public evaluation Models infer transformations from examples and solve unseen grid puzzles; the evaluation split is explicit ARC 97.9% #2 of 66+1 versions & slices semi-private Models infer transformations from examples and solve unseen grid puzzles; the evaluation split is explicit ARC 92.1% #4 of 69
ARC-AGI-3semi-private Measures interactive game solving through relative human action efficiency ARC 54.8% #1 of 19
Chess Puzzles Models choose the best next move from a novel chess position Epoch 74% #1 of 69
Math 14 Leads on 2
AA Analyst Agent Data-analyst tasks: the model explores data with tools and reports findings AA 51.3% #7 of 44
FrontierMath Erdos Models solve expert-written advanced mathematics problems on the named tier and edition Epoch 2.9% #2 of 13
FrontierMath Tier 4 (v2) Models solve expert-written advanced mathematics problems on the named tier and edition Epoch 97.6% #2 of 55
FrontierMath Tiers 1–3 (v2) Models solve expert-written advanced mathematics problems on the named tier and edition Epoch 93.7% #1 of 70
MathArenaArXivMath · April 2026 Answer accuracy on the named mathematics competition under the recorded effort MathArena 90.8% #1 of 17+9 versions & slices ArXivMath · August 2026 Answer accuracy on the named mathematics competition under the recorded effort MathArena 88.6% #3 of 14ArXivMath · June 2026 Answer accuracy on the named mathematics competition under the recorded effort MathArena 94.4% #1 of 24ArXivMath · May 2026 Answer accuracy on the named mathematics competition under the recorded effort MathArena 95% #1 of 19BrokenArXiv · April 2026 Answer accuracy on the named mathematics competition under the recorded effort MathArena 97.5% #1 of 15BrokenArXiv · August 2026 Answer accuracy on the named mathematics competition under the recorded effort MathArena 81.9% #2 of 13BrokenArXiv · June 2026 Answer accuracy on the named mathematics competition under the recorded effort MathArena 99.1% #3 of 24BrokenArXiv · May 2026 Answer accuracy on the named mathematics competition under the recorded effort MathArena 91% #2 of 18ArXivLean · June 2026 Answer accuracy on the named mathematics competition under the recorded effort MathArena 65.2% #1 of 9ArXivLean · September 2026 Answer accuracy on the named mathematics competition under the recorded effort MathArena 37.7% #3 of 6
Science 5 Leads on 2
Humanity's Last Exam Very hard expert questions across many fields, written to stump today’s models AA 53.1% #8 of 370
CritPt (physics) Research-level physics problems that need many careful reasoning steps AA 28.9% #12 of 342
Vals Science Index Scientific problem solving across the index’s component evaluations Vals 65.1% #1 of 19
SciCode Writing code for real scientific computing problems (physics, biology, math) AA 55.4% #28 of 146
Terminal-Bench Science0.1 Agents complete expert-authored research workflows across five scientific domains in a terminal AA 63.3% #1 of 46
Coding 11
Terminal-Bench4.0 Agents use a terminal to complete executable coding, system administration and research tasks in the 4.0 suite 54% #9 of 138+1 versions & slices 2.1 Older, easier version of Terminal-Bench; Still widely reported AA 89.9% #2 of 161
DeepSWE1.1 Agents change repository code to resolve software engineering tasks under the stated DeepSWE edition 73.2% #3 of 44
FrontierCode1.1 Agents implement changes in real repositories, graded against maintainer-defined requirements AA 47% #5 of 39
ITBench SRE IT incident tasks: diagnose and fix broken systems like a site-reliability engineer AA 48.6% #8 of 38
WebDev Arena Human pairwise preferences in the stated webdev vote pool Arena 1,774 #2 of 97+2 versions & slices October 23, 2026 Human pairwise preferences in the stated webdev vote pool Arena 1,774 #2 of 98October 19, 2026 Human pairwise preferences in the stated webdev vote pool Arena 1,774 #2 of 97
SWE-bench Pro (v2 full) Agents resolve repository issues under the network-locked v2 protocol; full and 51-task Hard results are separate Scale 96.9% #5 of 16
SWE-bench Pro (v2 hard) Agents resolve repository issues under the network-locked v2 protocol; full and 51-task Hard results are separate Scale 90.2% #4 of 16
MirrorCode Agents reimplement software projects by observing their behavior Epoch 46.7% #5 of 10
Agents 10 Leads on 3
AutomationBench-AA Agents orchestrate SaaS workflows through APIs; objectives receive credit only when the task has no guardrail violation AA 66.6% #11 of 139
AA-Briefcase1.1 · Elo Agents produce professional deliverables; blind pairwise comparisons produce the source Elo rating AA 1,504 #19 of 146
GDPval-AA (work tasks)2.1 Real work tasks from many professions (reports, spreadsheets, analyses), graded against expert work AA 1,537 #26 of 177
AutomationBenchZapier run End-to-end completion of business workflows using real SaaS applications Zapier 37.1% #2 of 2
EBR Bench Agents learn to play Earthborne Rangers across repeated playthroughs Epoch 76.2% #1 of 24
Furniture Assembly Models inspect furniture assembly photos, identify incorrect builds and locate the mistaken step Epoch 80% #2 of 31
SpireClimb Agents play Slay the Spire under the named game and difficulty setting Epoch 25.7% #1 of 7
SpireClimb StS1 A20 Heart Agents play Slay the Spire under the named game and difficulty setting Epoch 16.7% #2 of 7
SpireClimb StS2 A10 Agents play Slay the Spire under the named game and difficulty setting Epoch 34.7% #1 of 7
τ³-BankingAA Agents retrieve banking knowledge and execute tools to resolve customer-support workflows AA 40% #16 of 137
Multimodal 3
MMMU-Pro (vision) College-level questions that require understanding images: charts, diagrams, photos AA 86.4% #2 of 136
Arena Vision Human pairwise preferences in the stated vision vote pool Arena 1,280 #25 of 88
Chartography Expert chart reasoning over professional scientific, medical, financial and engineering charts Surge 71% #2 of 37
Writing 20 Leads on 1
Arena Textcreative writing · October 22, 2026 Human pairwise preferences in the stated text vote pool Arena 1,421 #47 of 197+16 versions & slices creative writing Human pairwise preferences in the stated text vote pool Arena 1,450 #26 of 195overall Human pairwise preferences in the stated text vote pool Arena 1,477 #23 of 196Chinese Human pairwise preferences in the stated text vote pool Arena 1,489 #51 of 198English Human pairwise preferences in the stated text vote pool Arena 1,448 #55 of 205French Human pairwise preferences in the stated text vote pool Arena 1,425 #84 of 164German Human pairwise preferences in the stated text vote pool Arena 1,455 #33 of 167instruction following Human pairwise preferences in the stated text vote pool Arena 1,455 #33 of 205Japanese Human pairwise preferences in the stated text vote pool Arena 1,404 #46 of 152Korean Human pairwise preferences in the stated text vote pool Arena 1,427 #26 of 156non-English Human pairwise preferences in the stated text vote pool Arena 1,432 #44 of 205Polish Human pairwise preferences in the stated text vote pool Arena 1,427 #56 of 142Russian Human pairwise preferences in the stated text vote pool Arena 1,442 #44 of 203Spanish Human pairwise preferences in the stated text vote pool Arena 1,423 #72 of 168overall · October 15, 2026 Human pairwise preferences in the stated text vote pool Arena 1,443 #46 of 188creative writing · October 15, 2026 Human pairwise preferences in the stated text vote pool Arena 1,421 #44 of 187overall · October 22, 2026 Human pairwise preferences in the stated text vote pool Arena 1,442 #50 of 198
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 82.8% #6 of 96
EQ-Bench Creative Writingv3 Creative Writing v3 · Elo: Judge-rated creative writing under the v3 prompt set and pairwise ranking EQ-Bench 2,174 #1 of 94
Lech Mazur Short-Story Judges compare stories written to ten required elements; the comparison score can be negative Lech 3.4 #4 of 45
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 8.6% #5 of 23
MLCR-AA Models synthesize long medical records; the headline credits concise answers judged both accurate and complete AA 32.8% #6 of 104
Vals MedCode Medical coding accuracy on patient encounters Vals 48.5% #25 of 78
Vals MedScribe Accuracy of clinical notes produced from consultation transcripts Vals 87.9% #15 of 80
Vals Legal Research Bench Grounded legal research; a task passes only if all criteria pass Vals 39.4% #27 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 338
GDP.pdfAA all-pass Models reason over professional PDF documents; an attempt passes only when every atomic criterion passes 31% #4 of 139
Arena Document Human pairwise preferences in the stated document vote pool Arena 1,468 #13 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 94.9% #2 of 346+1 versions & slices Epoch Models answer graduate-level science multiple-choice questions Epoch 95.8% #2 of 134
OTIS Mock AIME 20242025 Models solve the stated OTIS mock AIME mathematics problems Epoch 100% #1 of 131
No benchmark matches
CheaperGrok 4.6high+27%-14%$1.48Context: 1.05M → 500K
BetterGLM 5.3 FlashmaxOpen+18%-84%$0.28Context: 1.05M → 1M+5
CheaperQwen3.8 27BxhighOpen+20%-34%$1.15Context: 1.05M → 262K+5
Gemini 3.8 Flashmedium+14%-46%*$0.93Context: 1.05M → 1M+2
Kimi K3maxOpen+15%+16%$2.00Context: 1.05M → 1M+4