Google Cloud to Offer SandboxAQ's Large Quantitative AI Models for Drug Discovery via Marketplace
Key Insights
Google (search) has partnered with SandboxAQ (search) to make specialized scientific AI models available through Google Cloud Marketplace, targeting drug discovery and materials science.
SandboxAQ (search)'s Large Quantitative Models are trained on numerical data and scientific equations rather than text, making them suited for chemistry, biology, and physics applications.
The first biopharma-focused model, AQPotency, predicts molecules likely to bind to protein targets, a critical step in early-stage drug development.
Google (search), part of Alphabet Inc. (search), has announced a partnership with SandboxAQ (search) to bring specialized scientific AI models to the Google Cloud Marketplace, a move that aims to accelerate drug discovery, materials science, and semiconductor manufacturing by giving enterprise researchers and pharmaceutical labs access to powerful computational tools on demand.
The partnership marks a significant expansion of AI infrastructure for the life sciences, placing SandboxAQ (search)'s so-called Large Quantitative Models (LQMs) alongside Google (search)'s own Gemini language models in a hybrid workflow that researchers can rent through the cloud platform.
A Different Kind of AI for Science
Unlike widely known large language models such as Gemini, which are trained on text and excel at generating fluent prose, SandboxAQ (search)'s LQMs are trained on numerical data, scientific equations, and laboratory results. This distinction is central to their value proposition: large language models are "surprisingly unreliable at numbers," while quantitative models are designed to produce answers that are "a number or a structure, not a fluent paragraph," making them effective for chemistry, biology, and physics applications.
On Google (search) Cloud, researchers will be able to combine these quantitative models with Gemini, using the language model for reasoning and interface tasks while the quantitative model handles the underlying science. Google is also pairing the marketplace launch with "Gemini for Science," a bundle of tools that includes its AI co-scientist, the AlphaEvolve coding agent, an empirical research assistant, and NotebookLM, pitched as a way to speed up "the routine, laborious steps of the scientific method rather than to replace the scientist."
Biopharma Applications and Early Targets
The first model available on the marketplace will be AQCat, designed to identify promising catalysts and materials for semiconductors, batteries, and other technologies. The second model, AQPotency, is focused squarely on biopharma: it helps predict molecules with a high probability of binding to protein targets, a critical step in early-stage drug development.
SandboxAQ (search) CEO Jack Hidary said the technology can "save years of research while simultaneously expanding the solution space, opening up approaches that are unattainable with traditional methods." He noted that the company sees the greatest commercial potential today in biopharma, with clients already using the models to study glioblastoma (search), prostate cancer (search), Alzheimer's and Parkinson's diseases, and cardiovascular conditions (search).
Technical Infrastructure and Future Trajectory
Hidary also noted that while SandboxAQ (search)'s models use quantum algorithms, they currently run on traditional accelerators — GPUs and TPUs. As quantum computers develop, the computational power of these AI systems is expected to increase further.
SandboxAQ (search) began as a quantum lab within Alphabet's X division (formerly Google (search) X) and became an independent entity in 2022. Its investors include former Google CEO Eric Schmidt and Salesforce chief Marc Benioff. The company recently raised $500 million under the CHIPS government program to develop AI models for semiconductor manufacturing.
Strategic Context
The commercial logic behind the partnership is straightforward: Google (search) is competing with other hyperscalers to become the default infrastructure for enterprise AI, and scientific and industrial R&D represents a high-value segment that general-purpose chatbots do not serve well. Selling specialist models through the marketplace allows Google to capture that demand without having to build every domain model itself.
The partnership also strengthens SandboxAQ (search)'s position as a leader in scientific AI alongside Isomorphic Labs (search), DeepMind's drug-discovery spinoff that is moving toward clinical trials, and OpenAI (search), which recently released a model for biological research.
After the pandemic surge, the pharmaceutical market has faced falling demand and increased regulatory pressure. In this environment, speed and flexibility in development have become strategic advantages, and AI has become a key tool for acceleration. The marketplace listing represents a substantive shift: a category of AI that was largely confined to specialist labs becomes something a research team can rent, potentially democratizing access to advanced computational tools for drug discovery.