SandboxAQ brings its drug discovery models to Claude — no PhD in computing required

BitcoinWorld

SandboxAQ brings its drug discovery models to Claude — no PhD in computing required

Drug discovery remains one of the most expensive and failure-prone processes in modern industry. Finding a single viable molecule can take a decade and cost billions, and most candidates never reach the market. A wave of AI startups has promised to accelerate this pipeline, but many of their tools remain accessible only to researchers already comfortable with specialized computing infrastructure. SandboxAQ, a company spun out of Alphabet roughly five years ago, believes the real bottleneck isn’t the models themselves — it’s the interface.

Bridging the gap between scientific models and researchers

SandboxAQ has partnered with Anthropic to integrate its scientific AI models directly into Claude, the company’s conversational AI platform. The integration places powerful drug discovery and materials science tools behind a natural language interface, eliminating the need for users to set up their own computing environments. Nadia Harhen, SandboxAQ’s general manager of AI simulation, described the move as a first: a frontier quantitative model running on a frontier large language model accessible in plain language.

The company, chaired by former Google CEO Eric Schmidt, has raised more than $950 million from investors. Beyond drug discovery, SandboxAQ operates in cybersecurity and other quantitative fields. But its core differentiator lies in what it calls large quantitative models, or LQMs. These are physics-grounded models built on the rules of the physical world rather than patterns in text. They can run quantum chemistry calculations and simulate molecular dynamics and microkinetics — the step-by-step processes of chemical reactions at the molecular level.

Why physics-grounded models matter

Traditional AI models in drug discovery often rely on statistical correlations from existing data. SandboxAQ’s LQMs, by contrast, are trained on real-world lab data and scientific equations, allowing them to predict how candidate molecules will behave before any physical experiment begins. This approach can save pharmaceutical companies years of trial and error.

SandboxAQ’s typical customers include computational scientists, research scientists, and experimentalists at large pharmaceutical or industrial companies searching for new materials that can become marketable products. Harhen noted that these customers often come to SandboxAQ after trying other software that failed to translate computational results into real-world outcomes.

Implications for the broader AI economy

SandboxAQ frames its work within what it calls the quantitative economy — a $50+ trillion sector spanning biopharma, financial services, energy, and advanced materials. The company’s bet is that making quantitative models accessible through conversational AI will unlock value far beyond the current user base of computational specialists. While competitors like Chai Discovery and Isomorphic Labs focus on improving the science of the models themselves, SandboxAQ is betting that usability will be the deciding factor in real-world adoption.

Conclusion

SandboxAQ’s integration with Claude represents a practical step toward democratizing advanced scientific simulation. By removing the infrastructure barrier, the company hopes to accelerate drug discovery and materials development for organizations that lack deep computational resources. Whether this approach yields faster breakthroughs than model-centric competitors will depend on how effectively researchers adopt and trust the conversational interface for high-stakes scientific work.

FAQs

Q1: What are large quantitative models (LQMs)?
LQMs are AI models grounded in physics and real-world scientific data, designed to perform quantum chemistry calculations and simulate molecular dynamics. Unlike language models, they are built on the rules of the physical world.

Q2: How does the SandboxAQ-Claude integration work?
Users can interact with SandboxAQ’s LQMs through Anthropic’s Claude using natural language, without needing to set up their own computing infrastructure. The models run on Anthropic’s platform and respond to conversational queries.

Q3: Who is the target user for this tool?
Primarily computational scientists, research scientists, and experimentalists at pharmaceutical and industrial companies who need to simulate molecular behavior for drug discovery and materials development.

This post SandboxAQ brings its drug discovery models to Claude — no PhD in computing required first appeared on BitcoinWorld.