Anthropic says Claude Science accelerates biomolecular modeling with open-source optimizations

On November 7, 2026, Anthropic published new research claiming that Claude can substantially accelerate biomolecular modeling, a computational field used to predict and design molecular structures. In the research, the company reports that Claude Science helped optimize more than 30 open-source models for structure prediction and design, delivering an average speedup of roughly 4x with minimal loss of precision. Alongside the work, Anthropic says it open-sourced the optimization code and a low-memory mode, an addition aimed at teams constrained by RAM rather than GPU time. The results are framed as a step toward making structural biology simulations faster to run at scale, with the optimizations presented as broadly reusable rather than tied to a single pipeline. Anthropic is positioning the speedups alongside the open-source releases, presenting the gains as coming with little degradation in precision, though the company has not shared detailed benchmark comparisons beyond that average figure. As of this writing, the research has been published and the code made available, and no independent validation has been reported. Later coverage or verification steps have not appeared yet, so the current state of the story remains at the research release, with the open-sourced code and low-memory mode now public.