Anthropic Claude Fable 5.2 scores 43/43 on Plotly benchmark

Anthropic's Claude Fable 5.2 model has passed all 43 questions on Plotly's data analytics benchmark, becoming one of only three models to achieve a perfect score. Plotly reports that while the model successfully handled open-ended questions, hidden patterns, and statistics traps, it was also notably more expensive and slower per run compared to other top-scoring models like OpenAI's GPT-6.1 Sol. Open article →