turbopuffer
Free tierFast vector and full-text search engine built on object storage — 10x cheaper and infinitely scalable
Free tier available·Technical·API available
Key strengths
10x cheaper than traditional vector databases due to object storage architectureSub-10ms p50 latency with Memory/SSD caching layerMassive scale: 4T+ documents, 10M+ writes/s, 25k+ queries/s in productionHybrid search combining vector (ANN) and full-text (BM25) in one systemUnlimited namespaces with instant copy-on-write branching
Free tier + paid plans
San Francisco, USA
Founded 2023
No ratings yet
turbopuffer is a vector and full-text search database architected from first principles on object storage (S3-compatible). It uses a Memory/SSD cache tier to deliver sub-10ms p50 query latency while keeping storage costs dramatically lower than traditional in-memory or SSD-only vector databases. It supports approximate nearest neighbor (ANN) vector search, BM25 full-text search, hybrid search, and rich metadata filtering — all via a simple HTTP API. In production it handles 4T+ documents, 10M+ writes per second, and 25k+ queries per second across customers like Anthropic, Cursor, Notion, and Atlassian.