Claude Fable 5.1 Scores 73.4% on CursorBench 3.2

Melvin Vivas · X post · 2026-09-02 · Open on X

Topics: Industry Trends & Job Market, Evaluation (Evals) & Testing, AI Dev Tools & Productivity · Level: beginner

Summary

The creator shares a quote saying Claude Fable 5.1 is the most capable model run on CursorBench 3.2, with 73.4% at max effort. The quote says the model is especially good at checking its own work, which lets it finish hard coding tasks from start to end.

Key points

• CursorBench 3.2 score: 73.4% at max effort.
• Described as the most capable model run on that benchmark so far.
• Its main strength is verifying its own work, which supports end-to-end coding tasks.

Resources mentioned

• Claude Fable 5.1 · tool · anthropic.com · paid
  An Anthropic Claude model used as the performance reference for Opus 5.5.
  Also in: Claude Opus 5.5 Released: Fable 5.1-Level Performance at 40% Lower Cost (Melvin Vivas on X · notes), GPT-6 Astra Tops Vending-Bench, Beating Claude Fable 5.1 (Melvin Vivas on X · notes), Claude Fable 5.1 Runs a 38-Hour Unattended ML Task (Melvin Vivas on X · notes), Why Claude Fable 5.1 Costs Less: Cheaper Cache Reads (Melvin Vivas on X · notes)
• CursorBench · other · cursor.com · free
  Cursor's benchmark for comparing model performance on coding tasks.
  Also in: Claude Opus 5.5 in Cursor: top of CursorBench at 40% lower cost (Melvin Vivas on X · notes), Why Claude Fable 5.1 Costs Less: Cheaper Cache Reads (Melvin Vivas on X · notes)