Independent analysis of AI

Artificial Analysis Intelligence Index v4.3.2 incorporates 10 evaluations: AA-Briefcase v1.1, GDPval-AA v2.1, AutomationBench-AA, Terminal-Bench 4.0, SciCode, Humanity's Last Exam, GDP.pdf, CritPt, AA-Omniscience, AA-LCR v1.1

28 of 679 models

Add model from specific provider

###

Artificial Analysis Intelligence Index

Artificial Analysis Intelligence Index v4.3.2 includes: AA-Briefcase v1.1, GDPval-AA v2.1, AutomationBench-AA, Terminal-Bench 4.0, SciCode, Humanity's Last Exam, GDP.pdf, CritPt, AA-Omniscience, AA-LCR v1.1. See Intelligence Index methodology for further details, including a breakdown of each evaluation and how we run them.

Open Weights / ProprietaryReasoning / Non-ReasoningText Only / Multimodal InputsBy Country

Artificial Analysis Intelligence Index by Open Weights / Proprietary

28 of 679 models

Add model from specific provider

ProprietaryOpen WeightsOpen Weights (Commercial Use Restricted)

###

Artificial Analysis Intelligence Index

###

Open Weights

Indicates whether the model weights are available. Models are labelled as 'Commercial Use Restricted' if commercial use is limited by conditions, and as 'Non-commercial' if the license prohibits commercial use.

Cost per TaskTime per TaskOutput Tokens per Task

Cost per Intelligence Index Task

Weighted average cost (USD) per Artificial Analysis Intelligence Index task, segmented by token type. Lower is better

28 of 679 models

AnswerReasoningCache WriteCache HitInput

###

Cost per Intelligence Index Task

Weighted average cost per Intelligence Index task. Each evaluation’s cost is calculated from input, cache hit, cache write, reasoning, and answer token prices, divided by task count, and weighted by its Intelligence Index weight.

Intelligence Index vs. Cost per TaskIntelligence Index vs. Time per TaskIntelligence Index vs. Output Tokens per Task

Intelligence Index vs. Cost per Intelligence Index Task

Artificial Analysis Intelligence Index · Weighted average cost (USD) per Artificial Analysis Intelligence Index task

28 of 679 models

Most attractive quadrant

Pareto line

GoogleStepFunZ AIXiaomiMiniMaxOpenAIAnthropicMetaSpaceXAINVIDIAAlibabaDeepSeekMistralKimi

###

Cost per Intelligence Index Task

###

Artificial Analysis Intelligence Index

Intelligence Index vs. Cost per TaskIntelligence Index vs. Time per TaskIntelligence Index vs. Output Tokens per Task

Intelligence Index vs. Cost per Intelligence Index Task, by Model Release

All reasoning and effort variants of each selected release · Weighted average cost (USD) per Artificial Analysis Intelligence Index task

10 of 496 releases

Most attractive quadrant

Pareto line

AnthropicOpenAIMetaSpaceXAIXiaomiZ AIGoogleDeepSeekOther Models

###

Cost per Intelligence Index Task

###

Artificial Analysis Intelligence Index

Frontier Language Model Intelligence, Over Time

15 of 58 model creators

AnthropicOpenAIMetaSpaceXAIXiaomiAlibabaZ AIStepFunKimiGoogleDeepSeekInstitute of Foundation ModelsMiniMaxThinking MachinesMistral

###

Artificial Analysis Intelligence Index

Coding Agent IndexUpdated

Performance, cost, and execution time for leading coding agents on end-to-end software engineering tasks

IndexCostExecution Time

Artificial Analysis Coding Agent Index

Artificial Analysis Coding Agent Index v1.5 incorporates 3 benchmarks: DeepSWE v1.1, Terminal-Bench 4.0, and SWE-Atlas-QnA · Higher is better

Color by

ModelAgent

14 of 20 models

Artificial Analysis Coding Agent Index vs. Cost per Task

Artificial Analysis Coding Agent Index vs. average pay-per-token API cost per task (USD)

Color by

ModelAgent

14 of 20 models

Most attractive quadrant

Pareto line

AnthropicAnthropic + CognitionOpenAIOpenAI + CognitionSpaceXAIMetaZ.aiMoonshot AIAlibaba CloudDeepSeekGoogle

###

How to Read This Chart

Each point represents a coding-agent variant. Farther left means lower average cost per task, while higher on the chart means higher benchmark performance. The most efficient agents sit toward the upper-left: stronger results at lower cost.

CyberNew

Measuring model capability on enterprise cyber defense

Cyber IndexCyber Index vs. Cost per TaskCWE-Bench-AA vs. Cost per TaskDeepsecBench-AA vs. Cost per TaskCyberGym-E2E-AA vs. Cost per Task

Artificial Analysis Cyber Index

Artificial Analysis Cyber Index v1 incorporates 3 evaluations: CWE-Bench-AA, DeepsecBench-AA, and CyberGym-E2E-AA

Successes

Safety blocks

Image & Video

Top models from our Image Arena and Video Arena leaderboards, with 95% confidence intervals

Text to ImageImage EditingText to VideoImage to VideoVideo Editing

Text to Image Leaderboard

Elo scores from blind preference votes in our Image Arena. See the full leaderboard here.

15 of 163 models

Speech

Top models from our Text to Speech Arena, Speech to Text and Speech to Speech evaluations

Provider Voice ArenaControlled Voice ArenaAA-WER Index (Non-Streaming)AA-WER Streaming Index (Final Transcription)Speech to Speech Index

Provider Voice Arena Preference Elo

Arena Elo: average Elo rating of the model · Higher is better

15 of 93 models

###

Arena Preference Elo

Relative Elo score of the models as determined by responses from users in Artificial Analysis' Speech Arena. Some models may not be shown due to not yet having enough votes.

Capability Indexes

Measures the performance of models on specific capabilities and industries

Finance & AccountingStrategy & OpsLegalHealthcare & MedicalEngineeringEconomics

Artificial Analysis Finance & Accounting Index

Incorporates 7 evaluations: AA-Omniscience, GDPval-AA v2.1, AA-Briefcase v1.1, Humanity's Last Exam, AutomationBench-AA, AA-LCR v1.1, GDP.pdf · Higher is better

27 of 27 models

Benchmarks

Intelligence Evaluations

Intelligence evaluations measured independently by Artificial Analysis · Higher is better

CodingAgenticTool UsePrivate DatasetUser InteractionFinanceMedicalLegalIntelligence IndexLong ContextMultimodalInstruction FollowingFaithfulnessWritingBusinessSee more

19 of 27 evaluations

28 of 679 models

Add model from specific provider

AA-Briefcase v1.1Updated

Agentic knowledge work, (Elo-500)/2000

GDPval-AA v2.1Updated

Agentic real-world work tasks, (Elo-500)/2000

AutomationBench-AAUpdated

Agentic SaaS workflows

Terminal-Bench 4.0New

Agentic coding & terminal use

SciCode

Coding

Humanity's Last Exam

Reasoning & knowledge

GDP.pdfNew

Professional document reasoning, All-pass

CritPtUnder review

Physics reasoning

AA-Omniscience Accuracy

Knowledge

AA-Omniscience Non-Hallucination Rate

1 - hallucination rate

AA-LCR v1.1

Long context reasoning

Harvey LAB-AA

Legal agentic work, criterion pass rate

EnterpriseOps-Gym-AA

Agentic business operations

Terminal-Bench-Science 0.1New

Agentic scientific research workflows in a terminal

AA-AnalystAgent

Quantitative analysis on spreadsheets & documents

𝜏³-Banking

Agentic tool use

ITBench-AA

Kubernetes incident root-cause analysis

MMMU-Pro

Visual reasoning

MLCR-AA

Medical long context reasoning

###

Intelligence Evaluation Relevance

While model intelligence generally translates across use cases, specific evaluations may be more relevant for certain use cases.

###

Artificial Analysis Intelligence Index

AA-Briefcase v1.1Updated

AA-Briefcase is a frontier agentic evaluation for long-horizon knowledge work, testing agents on realistic business workflows that require deliverables such as spreadsheets, presentations, and memos

AA-Briefcase EloAA-Briefcase Rubric Score (%)Analytical Quality & Presentation EloAA-Briefcase Elo vs. Cost per Task

AA-Briefcase Elo

AA-Briefcase v1.1 is an agentic knowledge work benchmark developed by Artificial Analysis. AA-Briefcase Elo is a combined metric that aggregates rubric pass rate, analytical quality Elo and presentation Elo · Higher is better

27 of 203 models

Add model from specific provider

###

AA-Briefcase Elo

AA-Briefcase Elo is a combined metric that aggregates analytical quality Elo, presentation Elo, and rubric pass rate, with rubric performance converted into Elo via synthetic head-to-head matches. Elo and 95% confidence interval bounds are clamped at 0.

AA-Omniscience

AA-Omniscience is a knowledge and hallucination benchmark that rewards accuracy, punishes bad guesses and provides a comprehensive view of which models produce factually reliable outputs across different domains

AA-Omniscience IndexAA-Omniscience AccuracyAA-Omniscience Hallucination Rate

AA-Omniscience Index

AA-Omniscience Index (higher is better) measures knowledge reliability and hallucination. It rewards correct answers, penalizes hallucinations, and has no penalty for refusing to answer. Scores range from -100 to 100, where 0 means as many correct as incorrect answers, and negative scores mean more incorrect than correct.

27 of 554 models

Add model from specific provider

###

AA-Omniscience Index

GDPval-AA v2.1Updated

GDPval-AA v2.1 evaluates AI models on real-world, economically valuable tasks across a wide range of occupations

GDPval-AA v2.1 Leaderboard

Elo rating for performance on real-world work tasks · Anchored to DeepSeek V4.1 Flash (max) at 1,600 · Higher is better

27 of 261 models

Add model from specific provider

Openness Index

Artificial Analysis Openness Index assesses how 'open' models are on the basis of their availability and transparency across different components.

Openness Index ComponentsOpenness Index

Artificial Analysis Openness Index: Components

Openness Index underlying score contribution by components, up to a maximum of 18 (higher is more open)

10 of 321 models

Add model from specific provider

Transparency - Pre-training DataTransparency - Post-training DataTransparency - MethodologyModel Availability

Artificial Analysis Openness Index vs. Artificial Analysis Intelligence Index

10 of 321 models

Add model from specific provider

Most attractive quadrant

Pareto line

Z AIKimiDeepSeekAlibabaMiniMaxThinking MachinesNVIDIAMetaMistral

Output Tokens

Output tokens of leading AI models based on our independent evaluations

Output Tokens per TaskIntelligence Index vs. Output Tokens per TaskOutput TokensIntelligence Index vs. Output Tokens

Output Tokens per Intelligence Index Task

Weighted average number of output tokens used to run one task in the Artificial Analysis Intelligence Index

28 of 679 models

AnswerReasoning

###

Output Tokens per Intelligence Index Task

The number of tokens required per Intelligence Index task. This is calculated by multiplying the output tokens per eval by the relative weights of each benchmark in the Intelligence Index, then dividing by task count (excluding repeats).

Cost

Price and real-world costs of leading AI models based on our independent evaluations

Cost per TaskIntelligence Index vs. Cost per TaskEvaluation Breakdown

Cost per Intelligence Index Task

Weighted average cost (USD) per Artificial Analysis Intelligence Index task, segmented by token type. Lower is better

28 of 679 models

AnswerReasoningCache WriteCache HitInput

###

Cost per Intelligence Index Task

Total CostIntelligence Index vs. Total CostIntelligence Index vs. Compute

Cost to Run Artificial Analysis Intelligence Index

Cost (USD) to run all evaluations in the Artificial Analysis Intelligence Index

28 of 679 models

Add model from specific provider

OutputReasoningCache WriteCache ReadNon-Cache Input

###

Cost to Run Artificial Analysis Intelligence Index

The cost to run the evaluations in the Artificial Analysis Intelligence Index, calculated using the model's input, cache hit, cache write, reasoning, and answer token prices and the number of tokens used across evaluations (excluding repeats).

Cache Hit, Input, and Output PricingBlended PriceBlended Price (Stacked)Cache DiscountIntelligence Index vs. PriceIntelligence Index vs. Price (Log, Inverted)

Pricing: Cache Hit, Input, and Output

Price (USD per M Tokens)

28 of 679 models

Add model from specific provider

Cache HitInputOutput

###

Cache Hit

Price per token for cached prompts (previously processed), typically offering a significant discount compared to regular input price, represented as USD per million tokens. The values shown here are the cache hit price; cache write and cache storage are billed separately and vary by provider — see "Cache pricing by provider" for detail.

4 more notes

Speed & Latency

Comparison of first-party API performance

SpeedLatencyEnd-to-End Response TimeIntelligence Index vs. Speed

Output Speed

Output tokens per second · Higher is better

28 of 679 models

Add model from specific provider

###

Output Speed

Tokens per second received while the model is generating tokens (ie. after first chunk has been received from the API for models which support streaming).

###

Model Performance Representation

Figures represent performance of the model's first-party API or the median across providers where a first-party API is not available.

Time per TaskIntelligence Index vs. Time per TaskCost vs. Time per Task

Time per Intelligence Index Task

Weighted average decode time (minutes) per task; excludes TTFT and overhead time · Lower is better

28 of 679 models

###

Time per Intelligence Index Task

The weighted average time (seconds) per Artificial Analysis Intelligence Index task. This is calculated by dividing output tokens per task by output speed, weighted by the relative weights of each benchmark in the Intelligence Index.

Providers

Endpoint Accuracy Index: gpt-oss-120b (high)

v1.0 · Composite of BFCL v4-500, HLE-250 and AA-LCR-25 run against each provider endpoint · Percentage of the reference endpoint, with 95% confidence interval · Higher is better

16 of 16 providers

Reference (100%)Within referenceBelow reference

###

Endpoint Accuracy Index

Composite measure of how much of a model's accuracy a given provider endpoint preserves, from re-running BFCL v4-500, HLE-250 and AA-LCR-25 against that endpoint. Where a self-hosted reference endpoint exists, scores are expressed as a percentage of that reference (100 = matches reference); lower scores indicate accuracy lost to quantisation, sampling defaults, or other endpoint-side configuration. Scores are point-in-time snapshots. Methodology.

Output Speed vs. Price: gpt-oss-120b (high)

Output tokens per second · USD per 1M tokens (blended) · 10,000 input tokens

16 of 16 providers

Most attractive quadrant

Pareto line

AmazonAzureBasetenCerebrasCoreWeaveCrusoeDeepInfraDeepInfra (Turbo)Google VertexGroqNebius (Base)NovitaParasailSambaNovaScalewayTogether AI

###

Emerging Provider Competition

Smaller, emerging providers are offering high output speed and at competitive prices.

3 more notes

Pricing (Cache Hit, Input, and Output): gpt-oss-120b (high)

Price (USD per M Tokens) · Lower is better · 10,000 input tokens

16 of 16 providers

Cache HitInputOutput

###

Cache Hit

3 more notes

SpeedSpeed Over Time

Output Speed: gpt-oss-120b (high)

Output speed: output tokens per second · 10,000 input tokens

16 of 16 providers

###

Output Speed

Tokens per second received while the model is generating tokens (ie. after first chunk has been received from the API for models which support streaming).

###

Model Performance Representation

Figures represent performance of the model's first-party API or the median across providers where a first-party API is not available.

Independent analysis of AI

Independent analysis of AI

Independent analysis of AI

IntelligenceUpdated

Artificial Analysis Intelligence Index by Open Weights / Proprietary

Cost per Intelligence Index Task

Intelligence Index vs. Cost per Intelligence Index Task

Frontier Language Model Intelligence, Over Time

Coding Agent IndexUpdated

Artificial Analysis Coding Agent Index vs. Cost per Task

CyberNew

Artificial Analysis Cyber Index

Image & Video

Text to Image Leaderboard

Speech

Provider Voice Arena Preference Elo

Intelligence Evaluations

AA-Briefcase v1.1Updated

AA-Briefcase Elo

AA-Omniscience Index

GDPval-AA v2.1Updated

GDPval-AA v2.1 Leaderboard

Artificial Analysis Openness Index: Components

Artificial Analysis Openness Index vs. Artificial Analysis Intelligence Index

Output Tokens

Output Tokens per Intelligence Index Task

Cost

Cost per Intelligence Index Task

Cost to Run Artificial Analysis Intelligence Index

Pricing: Cache Hit, Input, and Output

Speed & Latency

Output Speed

Time per Intelligence Index Task

Providers

Endpoint Accuracy Index: gpt-oss-120b (high)

Output Speed vs. Price: gpt-oss-120b (high)

Pricing (Cache Hit, Input, and Output): gpt-oss-120b (high)

Output Speed: gpt-oss-120b (high)

Comments (0)