Benchling integrated NVIDIA BioNeMo models and NIM microservices into its Benchling AI platform, enabling protein structure prediction (OpenFold2 initially) and laying groundwork for generative modeling and property prediction inside Benchling workflows. The move aims to let scientists run biomolecular AI without switching tools or managing heavy infrastructure. Separately, Lila Sciences expanded its Series A with a $115 million Nvidia‑backed tranche to fund its vision of “scientific superintelligence” linking models to lab automation. Both announcements illustrate continued convergence of AI model providers, computational tooling and venture capital to accelerate molecular design at scale. Practical impact: Integrations that bring large biomolecular models into lab informatics can shorten design–build–test cycles, but adoption will hinge on model validation, data governance, and reproducible lab linkages.