Drug discovery remains one of the most expensive and failure-prone endeavors in modern industry. Bringing a single viable molecule to market can take a decade and cost billions, with most candidates never making it past clinical trials. A wave of AI startups has promised to change that, but many have only made the process less painful for researchers already comfortable with complex computational tools. SandboxAQ, an Alphabet spinout, believes the real bottleneck isn’t the underlying models — it’s the interface.
Making advanced science conversational
SandboxAQ has partnered with Anthropic to integrate its scientific AI models directly into Claude, the company’s conversational AI assistant. The integration places powerful drug discovery and materials science tools behind a natural-language interface, eliminating the need for specialized computing infrastructure. Users can now run quantum chemistry calculations and simulate molecular dynamics simply by asking.
Also read: 1win Crypto Tournaments Go Global With Up to 200K USDT in Rewards
Founded roughly five years ago as an Alphabet spinout, SandboxAQ counts Eric Schmidt, Google’s former CEO, as its chairman. The company has raised more than $950 million from investors and operates multiple business lines, including cybersecurity. But one of its most distinctive offerings is its large quantitative models, or LQMs.
These proprietary models are described as “physics-grounded,” meaning they are built on the rules of the physical world rather than patterns in text. They can simulate how chemical reactions unfold at the molecular level — a capability that tells researchers how candidate molecules are likely to behave before any lab work begins.
Also read: Apple’s Siri Revamp Could Introduce Auto-Deleting Chat Histories to Boost Privacy
Bridging the gap between science and usability
“For the first time, we have a frontier [quantitative] model on a frontier LLM that someone can access in natural language,” Nadia Harhen, SandboxAQ’s general manager of AI simulation, told TechCrunch. Previously, users of SandboxAQ’s LQMs had to provide their own digital infrastructure to run the models. The company’s customers are typically computational scientists, research scientists, or experimentalists working at large pharmaceutical or industrial companies searching for new materials that can become marketable products.
“Our customers come to us because they’ve tried all the other software out there, and the complexity of their problem is such that it didn’t work or didn’t yield positive results for them when that translation went to take place in the real world,” Harhen said.
The move contrasts with other well-funded AI-driven drug discovery companies like Chai Discovery and Isomorphic Labs, which have focused primarily on improving the underlying models. SandboxAQ is betting that accessibility — not raw computational power — will determine how quickly these tools transform the industry.
Why this matters for the broader economy
SandboxAQ’s LQMs are designed for what the company calls the “quantitative economy,” a $50+ trillion sector spanning biopharma, financial services, energy, and advanced materials. The company’s press release makes clear it isn’t building another chatbot or code assistant — it’s targeting the industries that AI is supposed to fundamentally reshape.
By making these models available through a conversational interface, SandboxAQ lowers the barrier to entry for scientists who may not have deep computational expertise. This could accelerate the discovery of new drugs, advanced materials, and more efficient chemical processes, potentially reducing the time and cost of bringing innovations from the lab to the market.
Conclusion
SandboxAQ’s integration of its physics-grounded models into Claude represents a significant shift in how advanced scientific AI tools are delivered. Rather than requiring specialized infrastructure and expertise, the partnership puts powerful simulation capabilities into the hands of researchers through a familiar conversational interface. As the pharmaceutical and materials science industries continue to explore AI-driven approaches, the focus on usability over raw capability may prove to be a decisive factor in how quickly these technologies are adopted.
FAQs
Q1: What are large quantitative models (LQMs)?
LQMs are AI models built on the rules of the physical world, designed to run quantum chemistry calculations and simulate molecular dynamics and microkinetics. They are trained on real-world lab data and scientific equations, making them distinct from large language models that rely on text patterns.
Q2: How does the SandboxAQ-Claude integration work?
SandboxAQ’s LQMs are accessible through Anthropic’s Claude conversational interface. Users can ask natural-language questions about molecular behavior, and the models run simulations in the background, returning results in a readable format without requiring users to set up their own computing infrastructure.
Q3: Who benefits most from this technology?
The primary beneficiaries are computational scientists, research scientists, and experimentalists at large pharmaceutical and industrial companies who need to evaluate candidate molecules or materials before committing to costly lab work. The conversational interface also makes the tools accessible to researchers without deep computational expertise.
Be the first to comment