Claude Fable 5.1 released with major coding and science gains

minchoi · x · 2026-09-02

Anthropic officially released Claude Fable 5.1, showing significant improvements over Fable 5 across benchmarks. Terminal-Bench rose from 42.0% to 55.8%, Science scores more than doubled from 24.7% to 52.6%, and CursorBench increased to 73.4%. The model offers better coding, long-running agent capabilities, and lower real-world costs.

Related event: Anthropic Launches Claude Fable 5.1 and Mythos 5.1 with 75% Cheaper Caching(56 posts)→

Original post →

More from Models

Models channel →