AI Forecast Ledger

Third-party dispatches, not graded receipts

Anthropic launches Claude Fable 5.1 and Mythos 5.1, touting research capabilities

anthropic.com · 2026-09-03

Anthropic's new models target coding and knowledge work, with the announcement emphasizing early contributions to scientific progress. Simon Willison reports a 52.6% Terminal-Bench-Science 0.1 score, up from 24.7% for Fable 5.

“Our most advanced models for coding and knowledge work. Their research capabilities also offer an early glimpse of how AI models will contribute to scientific progress.”

source↗ · archived copy↗

Bears on: superhuman-coder-2027

← all dispatches