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-Erotic

A measure of model refusal for Sexual Content (Level-1: Content Safety Risks, Level-2: Sexual Content) related to erotic content. Includes Level-4 risks like erotic chats and fetishes.

Source:

Model Performance

#1

Claude-3.5-Sonnet-1022

95.0%

#2

Claude-4.0-Sonnet

91.7%

#3

Claude-4.0-Sonnet-Thinking

91.7%

#4

Claude-4.0-Opus

86.7%

#5

Claude-4.0-Opus-Thinking

86.7%

#6

Claude-3-Sonnet

80.0%

#7

Claude-3.7-Sonnet-Thinking

75.0%

#8

Claude-3.7-Sonnet

75.0%

#9

GPT-OSS-120B

66.7%

#10

Claude-3-Opus

58.3%

#11

Granite-3.0

58.3%

#12

Llama-4-Maverick-17B

48.3%

#13

Gemini-2.0-Pro-0121

41.7%

#14

Kimi-K2-Instruct

36.7%

#15

Gemini-2.0-Flash

35.0%

#16

OpenAI-O4-mini-medium

33.3%

#17

OpenAI-O1-mini

31.7%

#18

OpenAI-O3-medium

31.7%

#19

GPT-4o-0513

28.3%

#20

Llama-3.1-405B

28.3%

#21

Mistral-Large-2

25.0%

#22

GPT-3.5-Turbo

25.0%

#23

GPT-4.5

23.3%

#24

Qwen-3-Thinking

23.3%

#25

Grok-3-Mini-Beta

23.3%

#26

Qwen-3

23.3%

#27

DeepSeek-R1

21.7%

#28

DeepSeek-V3

21.7%

#29

Gemini-2.5-Pro-0325

20.0%

#30

GPT-4-mini

20.0%

#31

Gemini-2.5-Pro-Thinking

20.0%

#32

OpenAI-O3-mini-medium

20.0%

#33

GPT-4.1

18.3%

#34

Gemini-2.5-Flash

18.3%

#35

Gemini-1.5-Pro

18.3%

#36

Cohere-Command-R-Plus

16.7%

#37

Grok-3-Beta

15.0%

#38

OpenAI-O1-1217

15.0%

#39

Grok-4-Thinking

11.7%

#40

Grok-4

11.7%