Nvidia just gave its AI biology platform a brain upgrade. At the BIO International Convention on June 23, the company announced the BioNeMo Agent Toolkit, a new framework that pairs its existing BioNeMo platform with agentic AI capabilities designed to automate and accelerate biological research workflows.

The toolkit creates specialized agents capable of tool invocation, interpreting scientific results, and speeding up hypothesis generation. Instead of researchers manually querying models and piecing together outputs, these agents can chain together multiple steps of a discovery workflow on their own.

What the toolkit actually does

The BioNeMo Agent Toolkit sits at the intersection of two existing Nvidia ecosystems. On one side, you have BioNeMo, Nvidia’s platform for AI-driven biology that includes open models and microservices tailored for drug discovery. On the other, there’s the broader NeMo Agent Toolkit, which Nvidia originally introduced around March 2025 as a general framework for building production-grade AI agents.

Early adopters are already building on it. Simulations Plus, a company focused on modeling and simulation for pharmaceutical research, is among the first partners. HighRes, with its Cellario OS laboratory automation platform, is another. Both are integrating the toolkit into their existing workflows.

The target audience is clear: diagnostics firms, pharmaceutical companies, AI-focused biotech startups, and the software developers building tools for all of them.

Building on a year of momentum

This announcement didn’t come out of nowhere. Back on January 12, 2026, Nvidia announced new AI models for the platform, including RNAPro, which focuses on RNA-related biological tasks. Then on May 19, 2026, the company revealed a collaboration with QIAGEN to integrate BioNeMo with biomedical knowledge graphs. Knowledge graphs are essentially structured databases of biological relationships, and plugging AI agents into them gives those agents a much richer understanding of the science they’re working with.

What this means for investors

The early partnership choices are worth watching closely. Simulations Plus (ticker: SLP) is a publicly traded company, and its integration with BioNeMo could serve as a bellwether for how the broader pharmaceutical software industry responds to agentic AI tools.

One risk worth flagging: AI agent technology is still relatively early-stage in terms of reliability for high-stakes scientific applications. Nvidia’s emphasis on leveraging the NeMo Agent Toolkit’s “production-oriented” agent development suggests the company is aware of this challenge.

The competitive landscape is also heating up. Google DeepMind, with AlphaFold and related tools, has established a strong presence in AI biology. Startups like Recursion Pharmaceuticals and Insilico Medicine are building their own AI-native drug discovery platforms.