An architect is designing a document analysis system that must process 150-page legal contracts (approximately 180,000 tokens) in a single pass, extracting key clauses and cross-referencing obligations across sections. Which Claude model configuration is most appropriate?

Claude 3 Sonnet or higher using the full 200K context window can process approximately 150,000-200,000 tokens in a single pass, making it ideal for long legal contracts where cross-referencing clauses requires global document awareness. Processing the entire contract in one call preserves context about how obligations interact across sections — critical for legal analysis. The map-reduce approach with 4,096-token chunks loses cross-document context and risks missing obligations that span pages. The sliding-window approach with overlapping 32K segments introduces complexity and still risks boundary artifacts at overlap points. The parallel-segment approach with 16 equal segments merged by a coordinator prompt breaks cross-clause relationships that legal analysis depends on.