Azure Cosmos DB AI Features

Azure Cosmos DB is Microsoft’s globally distributed, multi-model database with integrated vector search capabilities for AI applications. Supports document (NoSQL), MongoDB API, PostgreSQL API, Cassandra API, Gremlin (graph), and Table API, all on same underlying platform. Vector search (preview) enables similarity search within Cosmos DB NoSQL and MongoDB APIs without external vector database. DiskANN-based indexing provides high-performance approximate nearest neighbor search. Integrated with Azure OpenAI Service for embeddings generation via stored procedures. Pricing uses Request Units (RU/s) for throughput with serverless option for variable workloads. Global distribution enables single-digit millisecond reads worldwide with multi-region writes. Automatic indexing removes need for schema design, all properties indexed by default. Page should cover: API selection guide (when to use which API), vector search setup and limitations, Azure OpenAI integration pattern, pricing model explanation (RU/s), global distribution configuration, comparison with DynamoDB and MongoDB Atlas, and use cases for AI-powered applications.