You can now improve the relevance of any Cosmos DB query or search with the public preview of semantic reranker in Azure Cosmos DB. Semantic reranker analyzes the meaning of your queries and your data to reorder results based on contextual relevance rather than keyword matching alone. This ranking is especially useful for AI-powered applications such as retrieval-augmented generation (RAG), question answering, and agentic apps.
With semantic reranker, you can deliver more precise and reliable results without building complex ranking pipelines or tuning scoring logic manually. It works directly with your existing Cosmos DB queries—including vector, full-text, and hybrid search—enabling you to quickly enhance result quality while maintaining low operational overhead.