Drug discovery is one of the most expensive pursuits in modern industry, and finding a viable molecule may take ten years and cost billions of dollars, with most candidate drugs ultimately failing. Although many AI startups promise to improve this process, most tools still require highly skilled researchers to use them. However, SandboxAQ believes that the bottleneck is not in the model, but in the user interface.

SandboxAQ recently partnered with the AI company Anthropic, integrating its scientific AI model directly into the Claude platform. This means that powerful drug discovery and materials science tools can now be accessed through a conversational interface without requiring professional computing infrastructure.

Founded five years ago, SandboxAQ is a spin-off from Google, with Eric Schmidt, former CEO of Google, serving as its chairman. The company has raised over $950 million from investors and established multiple business lines, including a cybersecurity business. However, a unique aspect of SandboxAQ is that it develops large quantitative models (LQMs), which are proprietary models built based on physical laws rather than just text patterns. LQMs are capable of quantum chemistry calculations and simulate molecular dynamics and micro-dynamics — how chemical reactions unfold at the molecular level. This feature allows researchers to know the behavior of candidate molecules before laboratory experiments.

"LQMs are trained on real laboratory data and scientific equations, designed for a quantitative economy worth more than $50 trillion, covering areas such as biopharmaceuticals, financial services, energy, and advanced materials," the company said in a press release, clearly indicating that SandboxAQ is not developing another chatbot or code assistant, but pursuing the transformative economic impact that AI should have.

Different from Chai Discovery and Isomorphic Labs, which focus on science, SandboxAQ focuses on who can truly use these tools. "For the first time, we have a quantitative model on an advanced large language model that can be accessed through natural language," said Nadia Hagen, general manager of SandboxAQ's AI simulation department. Previously, users of SandboxAQ's LQMs had to provide their own digital infrastructure to run these models.

SandboxAQ's clients are usually computational scientists, research scientists, or experimental scientists, typically working in large pharmaceutical or industrial companies searching for new materials that can be turned into marketable products. "Our customers choose us because they have tried other software, but due to the complexity of the problems, these software could not successfully translate into real-world results," Hagen added.

Key Points:

🔬 SandboxAQ has partnered with Anthropic to integrate drug discovery and materials science tools into the Claude platform, simplifying the user interface.

💡 The company focuses on building quantitative models based on physical laws, simulating molecular behavior before experiments.

🚀 SandboxAQ aims to make its advanced AI technology accessible to more users, accelerating the drug discovery process.