Atlas

Benchmarks

LLM benchmarks

Browse 2,914 AI benchmarks with model scores and general-capability loadings.

BenchmarkCategoryModelsLeading score
CapabilityIndexes891169.9
MMLU ProGeneral QA49693.4
Humanity's Last Exam (Text-Only)General QA48863.1
Artificial Analysis Progress Index v1Indexes484351
GPQA Diamond (Artificial Analysis avg@5)General QA45696.3
SciCode (Artificial Analysis avg@3)Science41160.2
AA-LCR v1.1Search40688.7
GPQA DiamondGeneral QA40696.4
CritPt (avg@5)Science40232.9
AA-Omniscience IndexIndexes39146.4
Artificial Analysis Omniscience AccuracyGeneral QA39166.2
Artificial Analysis Omniscience Hallucination RateSafety3910.9
AA-LCRSearch35778.2
τ²-bench (Telecom)Agents33499.3
IFBench (Artificial Analysis avg@5)Chat & Writing33383.4
Terminal-Bench HardCode31265.9
MMLUGeneral QA29693.4
BuseyBench SVGMultimodal2828.0
LiveCodeBench (Artificial Analysis)Code27791.7
SnakeBench Average ApplesGames27519.8
SnakeBench Best ApplesGames27534.0
SnakeBench RatingGames27538.2
SnakeBench Total ApplesGames2751316
SnakeBench Win RateGames275100.0
Artificial Analysis Openness IndexIndexes26688.9
HumanEvalCode25899.3
Arena Text — English (No Style Control)Chat & Writing2531530
Arena Text — Exclude Ties (No Style Control)Chat & Writing2531569
Arena Text — Exclude Ties (Style Controlled)Chat & Writing2531554
Arena Text — Hard Prompts (No Style Control)Chat & Writing2531553
Arena Text — Hard Prompts English (No Style Control)Chat & Writing2531545
Arena Text — Instruction Following (No Style Control)Chat & Writing2531539
Arena Text — Non-English (No Style Control)Chat & Writing2531537
Arena Text — Non-English (Style Controlled)Chat & Writing2531521
Arena Text — Software & IT Services (No Style Control)Professional Work2531555
Arena Text — Writing, Literature & Language (No Style Control)Professional Work2531543
LMArena Text Arena EloChat & Writing2531528
LMArena Text Arena English EloChat & Writing2531523
LMArena Text Arena Hard Prompts EloChat & Writing2531548
LMArena Text Arena Hard Prompts English EloChat & Writing2531544
LMArena Text Arena Industry: Writing, Literature & Language EloChat & Writing2531529
LMArena Text Arena Instruction Following EloChat & Writing2531525
LMArena Text Arena Occupational Software & IT Services EloProfessional Work2531552
LMArena Text Overall No Style Control EloChat & Writing2531540
Arena Text — Creative Writing (No Style Control)Chat & Writing2521534
Arena Text — Entertainment, Sports & Media (No Style Control)Professional Work2521526
Arena Text — Life, Physical & Social Science (No Style Control)Professional Work2521546
Arena Text — Multi-turn (No Style Control)Chat & Writing2521565
LMArena Text Arena Coding EloCode2521573
LMArena Text Arena Creative Writing EloChat & Writing2521520
Loading Atlas data…