Pinecone

The most popular managed vector database, purpose-built for AI applications. Pinecone makes it simple to store and query high-dimensional embeddings at scale — powering the similarity search behind RAG systems, recommendation engines, and semantic search applications.

• Managed service — fully hosted vector database with zero infrastructure management • Fast queries — sub-50ms similarity search across billions of vectors • Serverless — pay-per-query pricing that scales automatically with traffic • Metadata filtering — combine vector search with attribute-based filters • Namespaces — organize vectors into logical partitions within an index • Integrations — native connectors for LangChain, LlamaIndex, and major AI frameworks

Free tier with 100K vectors. Serverless pricing from $0.33/1M read units.