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 6.5 #289 of 337
Knowledge 5
AA OmniscienceAccuracy Measures correct answers on the AA Omniscience factual knowledge test AA 15.2% #232 of 321+1 versions & slices Hallucination Rate Measures how often attempted answers are incorrect on the AA Omniscience factual knowledge test AA 79.4% #161 of 321
AA Omniscience Index Scores factual knowledge while penalizing incorrect answers to encourage admitting uncertainty AA -52.2 #221 of 321
IFBench (instructions) Checks whether the model follows precise formatting and content instructions AA 39.5% #161 of 248
EQ-Bench3 Judge-rated responses to social and emotionally nuanced scenarios, scored through the source pairwise ranking EQ-Bench 608 #48 of 50
Reasoning 4
ARC-AGI-1public evaluation Models infer transformations from examples and solve unseen grid puzzles; the evaluation split is explicit ARC 2.4% #61 of 61+1 versions & slices semi-private Models infer transformations from examples and solve unseen grid puzzles; the evaluation split is explicit ARC 0.5% #64 of 64
ARC-AGI-2public evaluation Models infer transformations from examples and solve unseen grid puzzles; the evaluation split is explicit ARC 0% #54 of 63+1 versions & slices semi-private Models infer transformations from examples and solve unseen grid puzzles; the evaluation split is explicit ARC 0% #59 of 66
Math 2
FrontierMath Models solve expert-written advanced mathematics problems on the named tier and edition Epoch 0% #40 of 40
MATH Level5 Models solve competition mathematics problems from the hardest MATH difficulty level Epoch 62.3% #18 of 35
Science 3
Humanity's Last Exam Very hard expert questions across many fields, written to stump today’s models AA 3.8% #318 of 337
CritPt (physics) Research-level physics problems that need many careful reasoning steps AA 0% #180 of 324
SciCode Writing code for real scientific computing problems (physics, biology, math) AA 21.3% #135 of 140
Coding 9
Terminal-Bench4.0 Agents use a terminal to complete executable coding, system administration and research tasks in the 4.0 suite 0% #75 of 132+1 versions & slices 2.1 Older, easier version of Terminal-Bench; Still widely reported AA 3.7% #140 of 156
SWE-rebenchall · 2025-01-01 to 2025-08-01 Repository issue resolution over the exact fresh pull-request window, under the source agent version Nebius 6.6% #5 of 6+5 versions & slices Go · 2025-01-01 to 2025-08-01 Repository issue resolution over the exact fresh pull-request window, under the source agent version Nebius 0% #1 of 6Java · 2025-01-01 to 2025-08-01 Repository issue resolution over the exact fresh pull-request window, under the source agent version Nebius 0% #1 of 6Python · 2025-01-01 to 2025-08-01 Repository issue resolution over the exact fresh pull-request window, under the source agent version Nebius 6.6% #5 of 6Rust · 2025-01-01 to 2025-08-01 Repository issue resolution over the exact fresh pull-request window, under the source agent version Nebius 0% #1 of 6TypeScript · 2025-01-01 to 2025-08-01 Repository issue resolution over the exact fresh pull-request window, under the source agent version Nebius 0% #1 of 6
Terminal-Bench Hard Agents solve difficult software, system-administration and data-processing tasks in a terminal AA 1.5% #197 of 242
Agents 4
AutomationBench-AA Agents orchestrate SaaS workflows through APIs; objectives receive credit only when the task has no guardrail violation AA 0.2% #131 of 133
AA-Briefcase1.1 · Elo Agents produce professional deliverables; blind pairwise comparisons produce the source Elo rating AA 0 #129 of 140
GDPval-AA (work tasks)2.1 Real work tasks from many professions (reports, spreadsheets, analyses), graded against expert work AA 0 #122 of 175
τ³-BankingAA Agents retrieve banking knowledge and execute tools to resolve customer-support workflows AA 3.3% #131 of 134
Multimodal 3
MMMU-Pro (vision) College-level questions that require understanding images: charts, diagrams, photos AA 52.9% #109 of 127
MMMUvalidation Image-grounded university-level questions across 30 disciplines; validation and test sets remain separate MMMU 69.4% #15 of 41
Arena Vision Human pairwise preferences in the stated vision vote pool Arena 1,118 #78 of 88
Writing 17
Arena Textcreative writing Human pairwise preferences in the stated text vote pool Arena 1,289 #138 of 195+14 versions & slices overall Human pairwise preferences in the stated text vote pool Arena 1,322 #143 of 196Chinese Human pairwise preferences in the stated text vote pool Arena 1,254 #147 of 181English Human pairwise preferences in the stated text vote pool Arena 1,304 #145 of 188French Human pairwise preferences in the stated text vote pool Arena 1,282 #122 of 149German Human pairwise preferences in the stated text vote pool Arena 1,273 #119 of 154instruction following Human pairwise preferences in the stated text vote pool Arena 1,248 #151 of 188Japanese Human pairwise preferences in the stated text vote pool Arena 1,206 #108 of 138Korean Human pairwise preferences in the stated text vote pool Arena 1,207 #117 of 144non-English Human pairwise preferences in the stated text vote pool Arena 1,252 #151 of 188Polish Human pairwise preferences in the stated text vote pool Arena 1,265 #120 of 133Russian Human pairwise preferences in the stated text vote pool Arena 1,263 #145 of 186Spanish Human pairwise preferences in the stated text vote pool Arena 1,278 #130 of 155overall · full cohort Human pairwise preferences in the stated text vote pool Arena 1,279 #150 of 188creative writing · full cohort Human pairwise preferences in the stated text vote pool Arena 1,249 #148 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 31.1% #90 of 95
EQ-Bench Creative Writingv3 Creative Writing v3 · Elo: Judge-rated creative writing under the v3 prompt set and pairwise ranking EQ-Bench 783 #84 of 91
Health & law 3
Vals LegalBench Legal reasoning and legal-language understanding on LegalBench Vals 72% #79 of 89
Vals MedCode Medical coding accuracy on patient encounters Vals 23.3% #75 of 78
Vals MedScribe Accuracy of clinical notes produced from consultation transcripts Vals 50.6% #79 of 80
Documents 2
AA-LCR (long context)1.1 Questions that need reading and connecting facts across very long documents (around 100K tokens) AA 27.7% #224 of 275
GDP.pdfAA all-pass Models reason over professional PDF documents; an attempt passes only when every atomic criterion passes 0% #131 of 133
Legacy · saturated or no longer reported 7
AIME2025 · AA archived AA evaluates the 30 integer-answer problems from the 2025 AIME mathematics competition AA 14% #121 of 140
LiveCodeBenchAA archived AA evaluates solutions to newly collected competitive-programming problems AA 29.9% #121 of 153
GPQA Diamond Graduate-level biology, chemistry and physics questions that can’t be solved by a quick web search 58.7% #232 of 315+1 versions & slices Epoch Models answer graduate-level science multiple-choice questions Epoch 51.8% #101 of 133
BFCLv4 Tests function calling, multi-turn interactions, search and memory in the v4 suite Berkeley 28.1% #28 of 39
OTIS Mock AIME 20242025 Models solve the stated OTIS mock AIME mathematics problems Epoch 7.8% #110 of 130
τ²-benchAA Tests service agents that use tools while coordinating with a simulated user AA 15.5% #210 of 245
No benchmark matches
CheaperGPT OSS 120BhighOpen×11.4-60%$0.092Context: 10M → 131K+2
OpenQwen3.6 27Bunstated×11.7+800%$0.60 / $3.60Context: 10M → 262K
Gemini 3.1 Flash Litehigh×10.3+275%$0.06Context: 10M → 1M
Claude Haiku 5.5high×12.5+33%$0.079Context: 10M → 1M+6