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 21.4 #112 of 339
Knowledge 6
AA OmniscienceAccuracy Measures correct answers on the AA Omniscience factual knowledge test AA 30.8% #75 of 323+1 versions & slices Hallucination Rate Measures how often attempted answers are incorrect on the AA Omniscience factual knowledge test AA 82.7% #190 of 323
AA Omniscience Index Scores factual knowledge while penalizing incorrect answers to encourage admitting uncertainty AA -30.8 #154 of 323
IFBench (instructions) Checks whether the model follows precise formatting and content instructions AA 78.8% #8 of 248
EQ-Bench3 Judge-rated responses to social and emotionally nuanced scenarios, scored through the source pairwise ranking EQ-Bench 1,451 #26 of 53
Mystery Game Puzzles Models choose the best move in puzzle positions from an undisclosed game Epoch 18% #43 of 65
Math 11
MathArenaAIME 2026 Answer accuracy on the named mathematics competition under the recorded effort MathArena 94.2% #18 of 24+9 versions & slices HMMT February 2026 Answer accuracy on the named mathematics competition under the recorded effort MathArena 87.9% #16 of 24Apex Shortlist Answer accuracy on the named mathematics competition under the recorded effort MathArena 60.1% #19 of 30ArXivMath · December 2025 Answer accuracy on the named mathematics competition under the recorded effort MathArena 38.2% #10 of 13ArXivMath · January 2026 Answer accuracy on the named mathematics competition under the recorded effort MathArena 54.9% #11 of 18Kangaroo 2025 1-2 Answer accuracy on the named mathematics competition under the recorded effort MathArena 81.3% #7 of 17Kangaroo 2025 11-12 Answer accuracy on the named mathematics competition under the recorded effort MathArena 91.7% #10 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 75% #8 of 17USAMO 2026 Answer accuracy on the named mathematics competition under the recorded effort MathArena 36.3% #7 of 8
FrontierMath Tiers 1–3 (v2) Models solve expert-written advanced mathematics problems on the named tier and edition Epoch 31.2% #58 of 70
Science 3
Humanity's Last Exam Very hard expert questions across many fields, written to stump today’s models AA 29% #87 of 339
CritPt (physics) Research-level physics problems that need many careful reasoning steps AA 1.7% #114 of 326
SciCode Writing code for real scientific computing problems (physics, biology, math) AA 44.8% #77 of 142
Coding 7
Terminal-Bench4.0 Agents use a terminal to complete executable coding, system administration and research tasks in the 4.0 suite 0% #77 of 134+1 versions & slices 2.1 Older, easier version of Terminal-Bench; Still widely reported AA 51.3% #69 of 156
ITBench SRE IT incident tasks: diagnose and fix broken systems like a site-reliability engineer AA 34.1% #25 of 37
WebDev Arena Human pairwise preferences in the stated webdev vote pool Arena 1,400 #69 of 97+2 versions & slices October 23, 2026 Human pairwise preferences in the stated webdev vote pool Arena 1,400 #69 of 98October 19, 2026 Human pairwise preferences in the stated webdev vote pool Arena 1,400 #69 of 97
Terminal-Bench Hard Agents solve difficult software, system-administration and data-processing tasks in a terminal AA 40.9% #32 of 242
Agents 7
AutomationBench-AA Agents orchestrate SaaS workflows through APIs; objectives receive credit only when the task has no guardrail violation AA 6.9% #78 of 135
AA-Briefcase1.1 · Elo Agents produce professional deliverables; blind pairwise comparisons produce the source Elo rating AA 562 #94 of 142
APEX-Agents Agents complete long-horizon professional-services tasks across realistic applications AA 15.3% #20 of 29
EnterpriseOps-Gym-AA Agents complete stateful enterprise workflows in eight domains; final database state determines success AA 30.9% #34 of 43
GDPval-AA (work tasks)2.1 Real work tasks from many professions (reports, spreadsheets, analyses), graded against expert work AA 792 #85 of 177
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 9.8% #20 of 21
τ³-BankingAA Agents retrieve banking knowledge and execute tools to resolve customer-support workflows AA 13.4% #76 of 134
Multimodal 2
MMMU-Pro (vision) College-level questions that require understanding images: charts, diagrams, photos AA 77.3% #37 of 128
Arena Vision Human pairwise preferences in the stated vision vote pool Arena 1,262 #39 of 88
Writing 19
Arena Textcreative writing · October 22, 2026 Human pairwise preferences in the stated text vote pool Arena 1,402 #61 of 197+16 versions & slices creative writing Human pairwise preferences in the stated text vote pool Arena 1,406 #61 of 195overall Human pairwise preferences in the stated text vote pool Arena 1,442 #61 of 196Chinese Human pairwise preferences in the stated text vote pool Arena 1,499 #38 of 198English Human pairwise preferences in the stated text vote pool Arena 1,443 #60 of 205French Human pairwise preferences in the stated text vote pool Arena 1,462 #43 of 164German Human pairwise preferences in the stated text vote pool Arena 1,444 #43 of 167instruction following Human pairwise preferences in the stated text vote pool Arena 1,423 #63 of 205Japanese Human pairwise preferences in the stated text vote pool Arena 1,430 #30 of 152Korean Human pairwise preferences in the stated text vote pool Arena 1,385 #54 of 156non-English Human pairwise preferences in the stated text vote pool Arena 1,430 #50 of 205Polish Human pairwise preferences in the stated text vote pool Arena 1,448 #30 of 142Russian Human pairwise preferences in the stated text vote pool Arena 1,430 #57 of 203Spanish Human pairwise preferences in the stated text vote pool Arena 1,441 #58 of 168overall · October 15, 2026 Human pairwise preferences in the stated text vote pool Arena 1,438 #54 of 188creative writing · October 15, 2026 Human pairwise preferences in the stated text vote pool Arena 1,402 #59 of 187overall · October 22, 2026 Human pairwise preferences in the stated text vote pool Arena 1,438 #56 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 68.3% #42 of 96
EQ-Bench Creative Writingv3 Creative Writing v3 · Elo: Judge-rated creative writing under the v3 prompt set and pairwise ranking EQ-Bench 1,478 #54 of 94
Health & law 1
MLCR-AA Models synthesize long medical records; the headline credits concise answers judged both accurate and complete AA 8.9% #61 of 103
Documents 2
AA-LCR (long context)1.1 Questions that need reading and connecting facts across very long documents (around 100K tokens) AA 77.3% #78 of 317
GDP.pdfAA all-pass Models reason over professional PDF documents; an attempt passes only when every atomic criterion passes 12.2% #50 of 135
Legacy · saturated or no longer reported 14
GPQA Diamond Graduate-level biology, chemistry and physics questions that can’t be solved by a quick web search 89.3% #44 of 315+1 versions & slices Epoch Models answer graduate-level science multiple-choice questions Epoch 86.4% #52 of 134
τ²-benchairline · 1.0.1 τ-bench official · 2-airline-1.0.1: Customer-service agents converse with simulated users, retrieve domain knowledge and change state through tools Sierra 81.5% #5 of 6+4 versions & slices overall · 1.0.1 τ-bench official · 2-overall-1.0.1: Customer-service agents converse with simulated users, retrieve domain knowledge and change state through tools Sierra 87.9% #1 of 6retail · 1.0.1 τ-bench official · 2-retail-1.0.1: Customer-service agents converse with simulated users, retrieve domain knowledge and change state through tools Sierra 84.4% #1 of 6telecom · 1.0.1 τ-bench official · 2-telecom-1.0.1: Customer-service agents converse with simulated users, retrieve domain knowledge and change state through tools Sierra 97.8% #1 of 6AA Tests service agents that use tools while coordinating with a simulated user AA 95.6% #16 of 245
OTIS Mock AIME 20242025 Models solve the stated OTIS mock AIME mathematics problems Epoch 82.2% #68 of 131
SWE-rebenchall · February 1, 2026 to March 1, 2026 Repository issue resolution over the exact fresh pull-request window, under the source agent version Nebius 59.9% #2 of 5+5 versions & slices Go · February 1, 2026 to March 1, 2026 Repository issue resolution over the exact fresh pull-request window, under the source agent version Nebius 0% #1 of 5Java · February 1, 2026 to March 1, 2026 Repository issue resolution over the exact fresh pull-request window, under the source agent version Nebius 0% #1 of 5Python · February 1, 2026 to March 1, 2026 Repository issue resolution over the exact fresh pull-request window, under the source agent version Nebius 59.9% #2 of 5Rust · February 1, 2026 to March 1, 2026 Repository issue resolution over the exact fresh pull-request window, under the source agent version Nebius 0% #1 of 5TypeScript · February 1, 2026 to March 1, 2026 Repository issue resolution over the exact fresh pull-request window, under the source agent version Nebius 0% #1 of 5
No benchmark matches
CheaperDeepSeek V4.1 Flash 1017maxOpen×2.7-61%$0.26Memory (4-bit): 242 GB → 458 GB+5
BetterGPT-5.6 Lunamax+17%-67%$0.18
CheaperClaude Haiku 5.5max+67%-85%$0.21Remove sampling parameters+5
CheaperGemini 3 Flash Previewlow+44%-17%$0.50 / $3
GLM 5.3maxOpen+83%+59%$2.01Memory (4-bit): 242 GB → 452 GB+5
Nemotron 3 UltraunstatedOpen+11%-22%$0.60 / $2.40Memory (4-bit): 242 GB → 336 GB+1