couchbase-examples/vector-search-cookbook

Demonstrates semantic search through two distinct indexing approaches: Couchbase's native Search Vector Index and SQL++ queries with vector similarity functions. Supports eight embedding providers (OpenAI, Azure, Anthropic, Cohere, Hugging Face, Jina, Mistral, Voyage) integrated with LLMs for retrieval-augmented generation, plus a caching layer (`CouchbaseCache`) for repeated query optimization. Jupyter notebooks provide end-to-end implementations requiring Couchbase 7.6+ with Search Service enabled.