About the tool
Serverless vector and full-text search database built on object storage.
turbopuffer runs vector and full-text search directly from object storage with an NVMe cache, trading always-hot infrastructure for substantially lower cost at scale. It serves production search and RAG retrieval for high-volume AI products. Usage-priced managed service with per-document and per-query billing.
- Pricing
- Paid
- Category
- Memory & RAG
- Maker
- Not publicly supplied
- Provenance
- Indexed by Attest from public information
✓ Attested