Iambic launches a 41B drug-discovery model

Iambic launched Enchant v3, a 41-billion-parameter multimodal model trained on 4.5 trillion tokens, more than 6,000 molecular properties, and 16 biomedical modalities. Iambic’s release says the model runs multiple hypotheses in parallel and will support internal and partner programs; in a separate AbbVie collaboration announcement, the companies said the model will be used across immunology, neuroscience, and oncology, while no clinical outcome is promised. Drug-discovery teams should file this as a data-integration and candidate-ranking signal—not a lab replacement—and compare it with the archive’s wet-lab AI pilot analysis.