Claude Fable 5.1 (high) scores 92.3% on WeridML, a user reports
The user reports a new benchmark high, beating Fable 5 (max) by 0.4%, with new records on 3 of the 17 tasks.
TLDR
A user running WeridML tests reports a record 92.3% score for Claude Fable 5.1 (high). They say the benchmark is close to saturation and question how long they’ll keep testing frontier models on it. In their September 3 update, they said a Fable 5.1 (max) run was underway. They also say 93.9% is achievable, with a somewhat higher score probably possible.
Combined views
—
1 Source, first seen ago
