Artificial Analysis Openness Index

CompositeDemo Purposes Only

Category

Composite

Source

Artificial Analysis

evaluation of record

Models covered

none ingested

Data status

Demo Purposes Only

invented numbers

Top score

not measured

Top model

Updated

no ingest

Full results

artificialanalysis.ai

on the source

A composite measure providing an industry standard to communicate model openness for users and developers.

Leaderboard

The shape of the source chart, drawn on invented numbers. We hold no data for this evaluation.

Demo purposes only · invented numbers · not a measurement

1Claude Opus 5Adaptive Reasoning, Max EffortAnthropic
78.02GPT-5.6 SolmaxOpenAI
70.33Kimi K3Kimi
62.64Grok 4.5highSpaceXAI
57.85GLM-5.2maxZ AI
50.66Muse Spark 1.1xhighMeta
47.07Gemini 3.5 FlashhighGoogle
41.38Qwen3.7 MaxAlibaba
38.09MiniMax-M3MiniMax
33.710DeepSeek V4 ProReasoning, Max EffortDeepSeek
29.811Motif 3BetaMotif Technologies
27.212MiMo-V2.5-ProXiaomi
24.113Hy3Tencent
21.914Nex-N2-ProNex AGI
19.715InklingxhighThinking Machines
17.516Agnes 2.5 Pro AlphaSapiens AI
15.917JT-4.1 Flash 236B A21BChina Mobile
14.018Nemotron 3 Ultra 550B A55BReasoningNVIDIA
12.419Claude Opus 5Adaptive Reasoning, Xhigh EffortAnthropic
10.820Claude Fable 5Adaptive Reasoning, Max Effort, Opus 4.8 FallbackAnthropic
9.5

Model names are drawn from our own roster; the scores are invented and measure nothing. See the source for the real leaderboard.

Source

Plain explanation

What it measures, how to read the number, and what to watch out for.

Scores how open a model release actually is, rather than whether its developer chose to call it open source — downloadable weights, license terms, and what was published about training all count separately. Higher is better. Its usefulness is in separating two things that routinely get conflated: a model can have fully open weights and still score poorly because the license forbids commercial use. Read it as a description of one release, not of a lab, since the same developer often ships one permissive model and one closed one. The weighting between those factors is an editorial judgment, so the ranking carries more information than the exact number.

© 2026 NYSGPT2525 LLC