Turbopuffer

Fast serverless vector database for AI application search

Data & AnalyticsFreemium

Turbopuffer is a serverless vector database designed for high-performance, cost-efficient similarity search at scale. It uses object storage (S3) as its primary store with local caching for hot data, enabling both cost efficiency and fast query performance. Turbopuffer supports hybrid search combining vector similarity with attribute filtering, making it suitable for RAG applications, semantic search, and recommendation systems that need cost-effective vector storage.

Key Features

• ✓Serverless architecture
• ✓S3-backed storage
• ✓Hybrid search
• ✓Attribute filtering
• ✓Cost-efficient
• ✓High performance