This study provides a comprehensive analysis of digital speech, orofacial and linguistic features for the assessment of schizophrenia. We recorded audio and video from 94 people with schizophrenia (pSCZ) and 100 healthy controls (HC) and extracted features automatically. Clinical rating scales were administered to assess positive, negative, and cognitive symptoms. We show that pSCZ exhibit significant alterations in speech timing, orofacial dynamics, and lexical richness, as compared to HC. A multimodal classification approach achieved high accuracy (96% AUC, 87% UAR), with speech features contributing most to discrimination. Correlation analysis revealed that speech timing and lip velocity measures are correlated with blunted affect and alogia. Linguistic features correlate well with positive symptoms, particularly conceptual disorganization and excitement. Cognitive abilities are most strongly associated with speech timing and specific linguistic features.