Index data

Learn how indexing works in Pinecone: serverless indexes, schemas, namespaces, integrated embedding, and metadata filtering.

Model your data in Pinecone using documents with dense_vector, sparse_vector, full-text string, and metadata fields for efficient retrieval.

Learn what changes in API version 2026-07, whether your existing code is affected, and how to move to schema-based indexes.

Compare Pinecone search types and choose the right retrieval approach: full-text (BM25), semantic (dense vector), sparse-vector, and hybrid.

Search a Pinecone index of dense vectors to find semantically similar records using text or vector queries, top_k results, and nearest neighbor lookup.

Run sparse-vector search in Pinecone with custom sparse encoders like pinecone-sparse-english-v0 for token-weighted keyword retrieval.

Narrow Pinecone search results by adding metadata filter expressions to your query, using operators like $eq, $in, $gt, and $and for precise retrieval.

Improve retrieval quality by reranking initial search results with a hosted or external model to surface the most relevant matches for RAG.