Generative intelligence explores the chemical space of ten million catalysts.

Discovery of catalytic materials requires systematic exploration of vast chemical spaces. However, the scope of exploration that can be achieved using conventional theoretical and experimental methods is very limited. Herein, we present a scalable framework based on distributed generative transformers, which integrates a transformer-based generative model with a distributed, parallel generation-screening workflow. By coupling dimensionality reduction with a machine-learning-potential (MLP) model for performance prediction, we construct a catalyst structure library comprising over ten million candidates-expanding the accessible design space by two orders of magnitude relative to existing generative models. Leveraging a pretrained model followed by task-specific fine-tuning, the tailored conditional generation strategy achieves >90% validity for target adsorbates such as CH3, thereby enabling focused exploration of methane conversion catalysts. Machine learning-accelerated screening of this massive library efficiently identifies 26 known active catalysts and more than 1200 previously unreported candidates. Subsequent subgroup discovery (SGD) analysis reveals synergistic elemental effects that modulate the surface electronic structure, tuning CH3 binding energies into the optimal window for CH4 activation. This work establishes a generalizable paradigm that seamlessly bridges generative exploration with high-throughput performance mapping, dramatically accelerating catalyst discovery across unprecedented chemical spaces.