Latent Labs actively participates in biosafety and biosecurity discussions with government and regulatory authorities regarding responsible development and preventing potential misuse. Access is restricted in compliance with international sanctions lists.

Latent-Y is a force multiplier for drug discovery teams. It enables researchers to pursue more targets and strategies in parallel, while retaining full visibility and control over design decisions.

Latent-Y can run fully autonomously end-to-end, or incrementally with human review at each stage. Every design decision is logged with the agent’s reasoning, allowing scientists to evaluate and build on its work.

Latent-Y operates in the same environment as protein design experts, with access to bioinformatics tools, biological databases, and external publications. It can parse research work plans and scientific papers as input.

Latent-X2 is the frontier generative model that designs therapeutic-grade antibodies and peptides with drug-like developability. Latent-Y is an AI agent that uses Latent-X2 as part of a larger autonomous workflow – handling target analysis, epitope selection, candidate design, computational validation, and iteration, all with expert-level biological reasoning.

Typical applications include autonomous antibody and peptide design campaigns. The agent handles the full workflow from target analysis to lab-ready sequences.

Contact partnerships@latentlabs.com.

Latent-Y is available through the Latent Labs platform, with options including cloud-based access and private deployments for enterprise partners

In our validation studies, developability profiles emerged directly from the model without iteration, on par or better than approved therapeutics. Nonetheless, we expect some designs to require further development and optimization.

We tested antibody designs against 18 soluble proteins, achieving picomolar to nanomolar hits against half the targets. Macrocyclic peptides were validated against 2 targets including K-Ras.