Converge Bio secures $25M Series A for GenAI drug discovery lab

Converge Bio is a textbook example of how generative AI is moving from abstract hype to deep vertical integration in life sciences. A $25 million Series A led by Bessemer, on top of a prior seed round, signals strong investor conviction that generative models trained on DNA, RNA, proteins and small molecules can become a standard part of drug discovery workflows—not a side experiment. The company explicitly brands itself as a ‘generative lab’ complementing wet labs, which is a powerful framing for how AGI-era tools might plug into science.

Their platform combines generative models, predictive filters, and physics-based docking into end‑to‑end systems that slot into CRO and pharma pipelines. That system‑level view is important for the AGI race: the labs that win may not just have the smartest models, but the tightest feedback loops tethered to real experimental outcomes. If Converge can consistently deliver 4x yield gains or faster antibody discovery, those empirical results will create a moat that’s hard for general-purpose LLM vendors to cross.

This deal also highlights a broader trend: capital is increasingly flowing to domain‑specific GenAI platforms with their own proprietary data and safety regimes. That could fragment the frontier into many specialized ‘AGI-like’ systems optimized for particular scientific or industrial domains, rather than a single general AGI serving every use case.