Anthropic 的 Claude Fable 5.2 在 Artificial Analysis 智能指数得分 65，居榜首

Artificial AnalysisK Artificial AnalysisModels Coding Agents Image, Speech, Video Inference Leaderboards About AI Trends K Anthropic •Claude Fable 5.2max •Proprietary model CompareTry it out API Provider Benchmarks Model summary IntelligenceUpdated #1 / 74 65 Artificial Analysis Intelligence Index 4 out of 4 units for Intelligence. Performance N/A Time per Intelligence Index task (minutes) Unknown out of 4 units for Performance. Cost #73 / 73 In $8.00Out $40.00Cache discount 98% $26.54 Cost per Intelligence Index task 4 out of 4 units for Cost. Verbosity #72 / 73 Input tokens 28.3BOutput/input 0.97% 275M Output tokens from Intelligence Index 4 out of 4 units for Verbosity. Comparison Summary Claude Fable 5.2 (Max, Default Fallback) 在智能方面处于领先模型之列，但与价格相近的其他模型相比，它的费用尤其高昂。该模型支持文本和图像输入，输出文本，并拥有 1M tokens 的上下文窗口。 Claude Fable 5.2 (Max, Default Fallback) 在 Artificial Analysis Intelligence Index 上得分 65，远高于同类模型的平均水平（中位数：40）。在评估 Intelligence Index 时，它生成了 275M tokens，与 72.8M 的中位数相比，显得非常冗长。 Claude Fable 5.2 (Max, Default Fallback) 的定价为每 1M 输入 tokens $8.00（昂贵，中位数：$2.00），每 1M 输出 tokens $40.00（昂贵，中位数：$10.00）。平均而言，在 Intelligence Index 上评估 Claude Fable 5.2 (Max, Default Fallback) 每个任务花费 $26.54。 ReasoningYes本页展示的是该模型的推理版本。 Context window1M~1500 A4 pages of size 12 Arial font 237 models in this class Metrics are compared against models of the same class: Artificial Analysis Intelligence Index v5.0.1 incorporates 9 evaluations: AA-Briefcase v1.1, AutomationBench-AA, Terminal-Bench 4.0, FrontierCode 2.0, AA-Omniscience v1.1, GDP.pdf, HLE-Diamond, CritPt v2 (Hard Subset), and Terminal-Bench-Science 0.1 27 of 734 models Add model from specific provider No publicly available API Artificial Analysis Intelligence Index Artificial Analysis Intelligence Index v5.0.1 includes: AA-Briefcase v1.1, AutomationBench-AA, Terminal-Bench 4.0, FrontierCode 2.0, AA-Omniscience v1.1, GDP.pdf, HLE-Diamond, CritPt v2 (Hard Subset), Terminal-Bench-Science 0.1. See Intelligence Index methodology for further details, including a breakdown of each evaluation and how we run them. 27 of 734 models Add model from specific provider No publicly available API ProprietaryOpen Weights (Commercial Use Restricted)Open Weights Artificial Analysis Intelligence Index Artificial Analysis Intelligence Index v5.0.1 includes: AA-Briefcase v1.1, AutomationBench-AA, Terminal-Bench 4.0, FrontierCode 2.0, AA-Omniscience v1.1, GDP.pdf, HLE-Diamond, CritPt v2 (Hard Subset), Terminal-Bench-Science 0.1. See Intelligence Index methodology for further details, including a breakdown of each evaluation and how we run them. 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. 27 of 734 models Most attractive quadrant Pareto line XiaomiOpenAIAnthropicInclusionAISpaceXAIKimiReflectionGoogleAlibabaZ AITencentMetaNVIDIAThinking MachinesMistralDeepSeek 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. Artificial Analysis Intelligence Index Artificial Analysis Intelligence Index v5.0.1 includes: AA-Briefcase v1.1, AutomationBench-AA, Terminal-Bench 4.0, FrontierCode 2.0, AA-Omniscience v1.1, GDP.pdf, HLE-Diamond, CritPt v2 (Hard Subset), Terminal-Bench-Science 0.1. See Intelligence Index methodology for further details, including a breakdown of each evaluation and how we run them. Measures the performance of models on specific capabilities and industries Incorporates 5 evaluations: AA-Briefcase v1.1, AA-Omniscience v1.1, HLE-Diamond, AutomationBench-AA, GDP.pdf · Higher is better 27 of 128 models Add model from specific provider No publicly available API Benchmarks Evaluations Evaluations measured independently by Artificial Analysis · AA-PersonalityBench scores are percentile ranks from 1 to 100 CodingAgenticTool UsePrivate DatasetUser InteractionFinanceMedicalLegalIntelligence IndexPersonalityLong ContextMultimodalInstruction FollowingFaithfulnessWritingBusinessSee more 23 of 44 evaluations27 of 734 models Add model from specific provider No publicly available API AA-Briefcase v1.1 Agentic knowledge work, Elo AA-Omniscience v1.1 AccuracyUpdated Knowledge AA-Omniscience v1.1 Non-Hallucination RateUpdated 1 - hallucination rate AutomationBench-AA Agentic SaaS workflows Terminal-Bench-Science 0.1New Agentic scientific research workflows in a terminal FrontierCode 2.0New Agentic software engineering in open-source repositories GDP.pdf Professional document reasoning, All-pass Terminal-Bench 4.0 Agentic coding & terminal use CritPt v2 (Hard Subset) Physics reasoning AA-AnalystAgent Quantitative analysis on spreadsheets & documents Harvey LAB-AA v1.1Updated Legal agentic work, Hallucination-Gated All-Pass Rate EnterpriseOps-Gym-AA Agentic business operations GDPval-AA v2.1 Agentic real-world work tasks, Elo AA-3DNew Agentic interactive 3D experiences, Elo Code2Video-AANew Product and brand videos designed in code, Elo ITBench-AA Kubernetes incident root-cause analysis MLCR-AAUpdated Medical long context reasoning AA-PersonalityBench: Expressiveness ScoreNew Rich, varied, well-crafted language AA-PersonalityBench: Organization ScoreNew Well-structured, scannable, answer-first AA-PersonalityBench: Transparency ScoreNew Truthful, acknowledges uncertainty, non-sycophantic AA-PersonalityBench: Slop ScoreNew

（原文较长，此处译出前 6000 字，完整内容请点「打开原文」。）