SandboxAQ puts 20,000x catalyst screening in Claude

SandboxAQ made AQCat generally available through Claude Science, where researchers can screen catalyst candidates in natural language using a model trained on 13.5 million high-fidelity DFT calculations across 47,000 catalyst systems; the company claims speeds up to 20,000× faster than conventional calculations. The tool is designed to rank thousands of materials before teams spend on detailed modeling or lab work, but the speed claim does not replace experimental validation. Scientists should treat this as a front-end screening multiplier and keep the wet-lab evidence gate that the archive’s AI-scientist replication work argues is still the scarce step.