A Guth Labs publication ↗
Back to News

Shanghai AI Lab opens Sage S2.1 Flash for scientific workflows

Conceptual scientific research illustration combining molecular, protein and computational imagery.
Guth News · AI-generated conceptual illustration

Shanghai AI Lab opens Sage S2.1 Flash for scientific workflows

By guthlabs.newsroom · November 16, 2026

Shanghai Artificial Intelligence Laboratory has made Sage S2.1 Flash available with downloadable weights and an FP8 variant, adding an openly licensed option for research workflows that combine scientific data, coding and tool use. The lab’s official repository says both code and model weights use the MIT license.

The distinctive feature is scientific input support. Through Sage SciPerception, the model can work with more than ten input types, including molecular graphs, crystal structures and protein sequences and structures. The system maps those inputs into representations that can be processed alongside natural-language instructions.

The lab reports a 90.3 score on MLE-bench Lite, ahead of the five comparison models in its table. That is a developer-reported result, not an independent validation or evidence that the system can autonomously conduct reliable science.

Deployment also involves more than downloading a chatbot. Scientific use requires the appropriate encoders, projectors, preprocessing dependencies and task-specific components. For research teams, the opportunity is a customizable model that can inspect intermediate outputs and revise a tool-driven analysis. The practical question is whether that capability holds up on their own datasets and verification requirements.

Sources and citations

Revision history

Published .