S1-Omni is presented as a unified multimodal reasoning model for scientific understanding, prediction, and generation. It maps natural-language instructions and scientific objects, including CIF files, SMILES strings, protein sequences, spectra, and scientific images, into a shared representation space. The model uses scientific laws and expert knowledge during data construction and training, then applies task-specific decoders for property prediction, spectrum-to-molecule generation, protein site and structure prediction, and scientific image generation or editing. Its reported S1-Omni-Corpus covers 200 scientific tasks and millions of reasoning samples, with evaluation on more than 60 scientific benchmarks. The abstract claims it outperforms GPT-5.5 and Gemini-3.1-Pro on most benchmarks and matches or exceeds domain-specific models on several tasks.
No heat snapshots are available in the last 24 hours.