SynthStrategy: Extracting and Formalizing Latent Strategic Insights from LLMs in Organic Chemistry

Abstract Modern computer-assisted synthesis planning (CASP) systems show promise at generating chemically valid reaction steps but struggle to incorporate strategic considerations such as convergent assembly, protecting-group minimization, and optimal ring-forming sequences. We introduce a methodology that leverages Large Language Models (LLMs) to distill synthetic knowledge into code. Our system analyzes synthesis routes and translates strategic principles into executable Python functions representing diverse strategic and tactical rules, such as strategic functional group interconversions, protecting strategies and ring construction features. We validate the use of LLM-as-a-judge in this domain, demonstrating their efficacy in evaluating synthesis-specific code generation. By formalizing this knowledge as verifiable code rather than simple heuristics, we create testable, interpretable representations of synthetic strategy. We release the complete codebase and the USPTO-ST data set─synthesis routes annotated with strategic tags. This framework unlocks a novel capability for CASP: natural language-based route retrieval, achieving 74.5% Top-3 accuracy on our benchmark. We further validate our library through temporal analysis of historical trends and chemically intuitive route clustering that offers more granular partitioning than common previous methods. This work bridges the tactical-strategic divide in CASP, enabling specification, search, and evaluation of routes by strategic criteria rather than structure alone.