7 of Anthropic's last 8 frontier models never got public evals
taoroalin · x · 2026-11-06
A pointed reminder circulating on X: 7 of Anthropic's last 8 frontier models never received public evals, meaning published benchmark results cover only a fraction of the lab's internally deployed models. The claim adds fuel to ongoing debates about frontier lab evaluation transparency, echoing the recent METR pilot controversy.
Related event: Insider: 7 of Anthropic's Last 8 Frontier Models Never Publicly Benchmarked(2 posts)→