Drug discovery is one of the most costly failures in modern industry. Finding a single viable molecule can take a decade and cost billions of dollars, and most candidates have yet to be successful. A generation of AI startups promises to solve this problem, and most of them are making the problem less painful for researchers, who are already technologically sophisticated enough to use the tools.
However, SandboxAQ believes that the bottleneck is not the model. It’s an interface.
The company partnered with Anthropic to integrate scientific AI models directly into Claude, placing powerful drug discovery and materials science tools behind a conversational interface that does not require the use of dedicated computing infrastructure.
Founded about five years ago as an Alphabet spinout, SandboxAQ is chaired by former Google CEO Eric Schmidt. The company, which has raised more than $950 million from investors, has built out various business lines, including a cybersecurity business.
However, one of the more unique things that SandboxAQ does is generate large-scale quantitative models (LQMs). These unique models are “physics-based,” meaning they are built on the rules of the physical world rather than patterns in text. You can perform quantum chemistry calculations and simulate both molecular dynamics and microkinetics, the study of how chemical reactions unfold at the molecular level. This is important because it tells researchers how a candidate molecule is likely to behave before they even set foot in the lab.
“Trained on real-world experimental data and scientific equations, LQM is an AI model designed for the quantitative economy, a $50 trillion-plus sector spanning biopharmaceuticals, financial services, energy, and advanced materials,” the company said in a news release, strongly hinting that Sandbox AQ is not building another chatbot or code assistant, but instead chasing the economy that AI is supposed to transform.
Chai Discovery and Isomorphic Labs are both well-funded bets on better models and are focused on science. SandboxAQ is focused on who can actually use it.
“For the first time, we have a Frontier (quantitative) model on Frontier LLM that someone can access in natural language,” Nadia Harhen, general manager of AI simulation at SandboxAQ, told TechCrunch. Previously, users of SandboxAQ’s LQM had to provide their own digital infrastructure to run their models.
SandboxAQ’s customers tend to be computational scientists, research scientists, or experimenters. Typically, these people work for large pharmaceutical or industrial companies and are looking for new materials that can be turned into marketable products.
“Our customers come to us because they have tried all the other software out there, but the problem is so complex that when they actually try to translate it, it doesn’t work or they don’t get good results,” says Harhen.
If you buy through links in our articles, we may earn a small commission. This does not affect editorial independence.
