SandboxAQ’s AQCat now runs on Claude Science for materials discovery

SandboxAQ has launched AQCat, an AI model capable of screening potential catalysts up to 20,000 times faster than conventional laboratory methods that typically evaluate fewer than 100 materials each week. The tool, now generally available on Claude Science, predicts catalyst effectiveness by calculating adsorption energy, a key early indicator, while also accounting for magnetic behavior often ignored by other machine-learning models. “In catalysis research, a key bottleneck has been the number of surfaces you can realistically evaluate within compute and scientist resource constraints,” says Dr. Joe Gauthier of Texas Tech University; AQCat’s speed changes the scale of research questions and opens previously inaccessible design spaces. AQCat Leverages Spin-Aware Physics for Rapid Catalyst Screening AQCat’s ability to accurately calculate adsorption energy marks a departure from many existing machine-learning models, which often overlook crucial magnetic properties. As SandboxAQ describes it, this brings commonly available and inexpensive metals like iron, cobalt, and nickel into viable consideration for catalyst development. The model was trained on AQCat25, a publicly available dataset containing 13.5 million high-fidelity Density Functional Theory calculations covering 47,000 catalyst systems and all industrially relevant elements. This acceleration of research fundamentally alters the scale of inquiry possible for materials scientists, as Joe Gauthier, Assistant Professor of Chemical Engineering at Texas Tech University, notes that the increased speed “changes the scale of questions my group can ask, and it opens up design spaces that were previously out of reach.” SandboxAQ built AQCat to recover near-DFT accuracy for magnetically complex systems while drastically reducing computational demands. Aayush Singh, Head of Science, Catalysis at SandboxAQ, explained that the tool allows researchers to pre-screen large numbers of potential catalysts before committing to more intensive calculations or experiments, shortening innovation cycles that previously spanned years. Geoff Ling, Founding Director of the Biotech Office at DARPA, emphasized the broader impact, stating that “Anything that helps researchers reduce uncertainty is incredibly valuable, and SandboxAQ’s MCP tools in Claude do that for both materials science and drug discovery.” Model Context Protocol Enables AQCat Access via Claude Science SandboxAQ has broadened access to its catalyst screening model, AQCat, through integration with Claude Science via a Model Context Protocol. This pairing allows researchers to query the Large Quantitative Model, trained to predict material physics, using natural language, eliminating the need for specialized coding skills. According to Dr. Joe Gauthier of Texas Tech University, this expanded accessibility is expected to reshape research workflows. Source: https://www.sandboxaq.com/press/sandboxaq-makes-aqcat-generally-available-on-claude-for-high-throughput-catalyst-screening More like thisArtificial IntelligenceSandboxAQ’s Models Link Large Language AI to $50T EconomyArtificial IntelligenceAQCat Adsorption Spin Now Accessible Via Waitlist IntegrationAnalog ComputingAQCat25: 11M-Point Dataset for Catalyst DesignQuantum Computing Business NewsSandboxAQ raises $300 million for Quantum AI SolutionsStay currentSee today’s quantum computing news on Quantum Zeitgeist for the latest breakthroughs in qubits, hardware, algorithms, and industry deals. Tags: Dr. Donovan Dr. Donovan is a futurist and technology writer covering the quantum revolution. Where classical computers manipulate bits that are either on or off, quantum machines exploit superposition and entanglement to process information in ways that classical physics cannot. Dr. Donovan tracks the full quantum landscape: fault-tolerant computing, photonic and superconducting architectures, post-quantum cryptography, and the geopolitical race between nations and corporations to achieve quantum advantage. The decisions being made now, in research labs and government offices around the world, will determine who controls the most powerful computers ever built. Latest Posts by Dr. Donovan: Symmatrics launches a pledge for AI and quantum security September 1, 2026 IQM award goes to researcher studying qubit performance August 31, 2026 Quantum Zeitgeist Weekly Digest August 31, 2026