Using only clinical vignettes, LLMs generated focused differentials diagnosis for two confirmed cases, COL3A1-related connective tissue disorder and COL4A1-related cerebral small vessel disease, also directed attention toward targeted molecular testing. These findings align with benchmark studies demonstrating LLM diagnostic accuracy in rare disease vignettes. For four unresolved cases, outcomes illustrated a four-role framework: Diagnostic clarification with management impact — PACS1 splice variant reinterpretation enabled ASM adjustment, improving seizure control. Multidisciplinary synthesis for resolution — DES desminopathy identified through VUS reclassification and reanalysis of imaging/pathology after decades of inconclusive workups. Clinical reframing without single etiology — a complex multisystem case redirected care from repeated testing toward anticipatory planning. Highlighting uncertainty and limits — a tumefactive perivascular space case showed divergent outputs, underscoring the need for physician oversight.  Compatible outputs were observed between general-purpose LLMs (GPT) and a domain-specific model (OpenEvidence), supporting reproducibility across platforms.

Readily available open-access LLMs, with physician judicious guidance, have potential to improve diagnostic efficiency, accuracy, and care quality in rare diseases. Prospective, multi-center validation will be essential to establish reproducibility, safety, and patient-centered benefit.