Artificial Analysis Intelligence Index v5 adds private test sets and drops three benchmarks
Article timeline
The articles in this story. Times are ET.
🚨 AI News | TestingCatalogXArtificial Analysis releases Intelligence Index v5 with mostly private test setsAIArtificial Analysis has released Intelligence Index v5, with 60% of the index now based on private test sets or held-out answers. Claude Opus 5.5 (max) leads at 58, ahead of GPT-6 Astra (max, 57), while the top open-weights model, Kimi K3.1 Pro, scores 46 but has not released its weights.
Ethan MollickXArtificial Analysis finds open-weights models falling further behind closed modelsAIArtificial Analysis moves to closed benchmarks and drops GDPval-AA, finding open-weights models trail the frontier by more than earlier measurements showed. Its v5 index puts Kimi K3.1 Pro, whose weights are not yet released, at 46, 12 points ahead of the best released-weights models, DeepSeek V4.1 Flash 1017 (max) and GLM-5.3 (max), which score 32. The gap is widest on HLE-Diamond and Terminal-Bench-Science, where Kimi K3.1 Pro scores 37% and 34%, against 68% for Claude Opus 5.5 (max) and 63% for GPT-6 Astra (max).
Boris PowerXOpenAI models lead Artificial Analysis V5 price-per-intelligence plotAIBoris Power says the V5 AAI index shows OpenAI models leading on price per performance across the curve, except at the most expensive Pareto-optimal point. He calls the Terra and Sol models super cheap and amazing.
Artificial AnalysisOfficialArtificial Analysis releases Intelligence Index v5 with new benchmarks and private test setsAIArtificial Analysis has released Intelligence Index v5, its largest upgrade since v4 in January, with 60% of the index now based on private test sets or held-out answers. The update adds FrontierCode 2.0 and Terminal-Bench-Science, upgrades HLE-Diamond, AA-Omniscience v1.1, and CritPt v2, and removes GDPval-AA v2.1, SciCode, and AA-LCR v1.1. Claude Opus 5.5 (max) scores 58 at the top, and the best released open-weights model scores 32, 26 points behind.
- 🚨 AI News | TestingCatalogXArtificial Analysis teases upcoming Intelligence Index v5
AIArtificial Analysis says its Intelligence Index v5 is coming, but the index is not live yet. The X post teasing it asks whether the leaderboard will change much, and a reply from Artificial Analysis says the launch is tomorrow.
- Artificial Analysis ArticlesOfficialPickArtificial Analysis releases Intelligence Index v5 with new private benchmarks
AIArtificial Analysis has released Intelligence Index v5, its biggest update since v4, with 60% of the index now based on private test sets or held-out answers. Claude Opus 5.5 (max) leads at 58, ahead of GPT-6 Astra (max, 57), and the best open weights model, Kimi K3.1 Pro, scores 46 but has not released its weights. The update also adds FrontierCode 2.0 and Terminal-Bench-Science and launches the Progress Index v1 for tracking capability over time.