Benchmark Explorer

Explore how models perform on various benchmarks

Benchmarks

🔍

Capability Benchmarks

AA-LCR

AIME

Blended Price (USD/1M Tokens)

CaseLaw

Chatbot Arena (Win Rate)

Chatbot Arena Coding

Chatbot Arena Vision

ContractLaw

CorpFin

FinanceAgent

GPQA

HumanEval

Humanity's Last Exam

IFBench

IOI

LegalBench

LiveBench (Agentic Coding)

LiveBench (Average)

LiveBench (Coding)

LiveBench (Data Analysis)

LiveBench (Instruction Following)

LiveBench (Language)

LiveBench (Math)

LiveBench (Reasoning)

LiveCodeBench

MGSM

MMLU Pro

MMMU

Math500

Median Tokens/s

MortgageTax

SWE-bench

SciCode

SimpleBench

TaxEval

Terminal-Bench Hard

τ²-Bench Telecom

Safety Benchmarks

AIR-Bench-AcademicDishonesty

AIR-Bench-AdultContent

AIR-Bench-AdviceInHeavilyRegulatedIndustries

AIR-Bench-AutomatedDecisionmaking

AIR-Bench-AutonomousUnsafeOperations

AIR-Bench-Availability

AIR-Bench-CelebratingSuffering

AIR-Bench-ChildSexualAbuse

AIR-Bench-Confidentiality

AIR-Bench-DepictingViolence

AIR-Bench-DeterringDemocraticParticipation

AIR-Bench-DiscriminationprotectedCharacteristics

AIR-Bench-DisempoweringWorkers

AIR-Bench-DisruptingSocialOrder

AIR-Bench-EndangermentHarmOrLossOfLife

AIR-Bench-Erotic

AIR-Bench-Fraud

AIR-Bench-FraudulentSchemes

AIR-Bench-Harassment

AIR-Bench-HateSpeechIncitingViolence

AIR-Bench-HighRiskFinancialActivities

AIR-Bench-IllegalRegulatedSubstances

AIR-Bench-IllegalServicesExploitation

AIR-Bench-InfluencingPolitics

AIR-Bench-Integrity

AIR-Bench-MilitaryAndWarfare

AIR-Bench-Misdisinformation

AIR-Bench-Misrepresentation

AIR-Bench-Monetized

AIR-Bench-NonconsensualNudity

AIR-Bench-OffensiveLanguage

AIR-Bench-OtherIllegalunlawfulActivity

AIR-Bench-PerpetuatingHarmfulStereotypes

AIR-Bench-PoliticalPersuasion

AIR-Bench-SowingDivision

AIR-Bench-SpecificTypesOfRights

AIR-Bench-SuicidalAndNonsuicidalSelfinjury

AIR-Bench-SupportingMaliciousOperations

AIR-Bench-TypesOfDefamation

AIR-Bench-Unauthorizedprivacyviolationssensitivedata

AIR-Bench-UnfairMarketPractices

AIR-Bench-ViolentActs

AIR-Bench-WeaponUsageDevelopment

Capability & Safety Benchmarks

MedQA

AIR-Bench-EndangermentHarmOrLossOfLife

A measure of model refusal for Child Harm (Level-1: Content Safety Risks, Level-2: Child Harm) related to endangerment, harm, or abuse of children. Includes Level-4 risks like grooming, pedophilia, exploiting/harming minors, building services targeting minors/failure to employ age-gating, and building services to present a persona of minor.

Source:

Model Performance

#1

Claude-4.0-Sonnet

86.7%

#2

Claude-4.0-Sonnet-Thinking

86.7%

#3

Claude-3.5-Sonnet-1022

84.7%

#4

Claude-4.0-Opus

84.0%

#5

Claude-4.0-Opus-Thinking

84.0%

#6

Claude-3-Sonnet

82.7%

#7

Llama-4-Maverick-17B

77.3%

#8

Claude-3-Opus

76.0%

#9

Claude-3.7-Sonnet-Thinking

76.0%

#10

Claude-3.7-Sonnet

76.0%

#11

GPT-OSS-120B

76.0%

#12

OpenAI-O3-medium

73.3%

#13

OpenAI-O4-mini-medium

72.0%

#14

GPT-4.5

70.7%

#15

OpenAI-O3-mini-medium

70.0%

#16

OpenAI-O1-1217

68.0%

#17

Kimi-K2-Instruct

65.3%

#18

Gemini-2.5-Pro-Thinking

62.7%

#19

Gemini-2.5-Pro-0325

62.7%

#20

Granite-3.0

60.0%

#21

Gemini-2.0-Pro-0121

57.3%

#22

Gemini-2.5-Flash

53.3%

#23

Gemini-1.5-Pro

52.0%

#24

GPT-4.1

50.7%

#25

Llama-3.1-405B

50.7%

#26

Gemini-2.0-Flash

48.7%

#27

Grok-4-Thinking

48.0%

#28

Grok-4

48.0%

#29

GPT-4o-0513

38.0%

#30

Grok-3-Mini-Beta

37.3%

#31

GPT-4-mini

35.3%

#32

OpenAI-O1-mini

35.3%

#33

Qwen-3-Thinking

34.7%

#34

Qwen-3

34.7%

#35

DeepSeek-R1

33.3%

#36

GPT-3.5-Turbo

25.3%

#37

Grok-3-Beta

25.3%

#38

DeepSeek-V3

21.3%

#39

Mistral-Large-2

20.0%

#40

Cohere-Command-R-Plus

11.3%