DeepCyte launched with $1.5 million in seed funding, positioning its single-cell metabolomics platform and AI software to predict drug toxicity and mechanisms at single-cell resolution. The company released MetaCore for high-throughput single-cell metabolomics and DeeImmuno, an AI model trained on proprietary atlases to classify toxicity and infer mechanisms. DeepCyte said its evaluation across 100 drugs enabled predictions of 17 toxicity mechanisms with 94% accuracy. By focusing on human cellular heterogeneity, the startup aims to help biopharma teams de-risk candidate selection earlier in development. For the market, the seed launch adds to growing interest in single-cell phenotyping tools that can connect preclinical biology to human safety outcomes.