turbopuffer - fast search engine built on object storage

# turbopuffer docs > turbopuffer is a vector and full-text search engine > built from first principles on object storage: fast, 10x cheaper, > and extremely scalable. This is the complete documentation for > the turbopuffer API and platform. > In production: 4T+ documents, 25k+ queries/s, 10M+ writes/s. - [Introduction](https://turbopuffer.com/docs/index.md): What turbopuffer is and how it works - [Quickstart](https://turbopuffer.com/docs/quickstart.md): Get started with the turbopuffer API using code examples - [Roadmap & Changelog](https://turbopuffer.com/docs/roadmap.md): Upcoming features and detailed changelog ## Database - [Architecture](https://turbopuffer.com/docs/architecture.md): System design: object storage, caching, WAL, and query routing - [Regions](https://turbopuffer.com/docs/regions.md): Available cloud regions and cross-region deployment options - [Concepts](https://turbopuffer.com/docs/concepts.md): turbopuffer's concepts - [Guarantees](https://turbopuffer.com/docs/guarantees.md): Guarantees of turbopuffer's architecture - [Tradeoffs](https://turbopuffer.com/docs/tradeoffs.md): Tradeoffs of turbopuffer's architecture - [Limits](https://turbopuffer.com/docs/limits.md): Current limits for turbopuffer. - [Branching](https://turbopuffer.com/docs/branching.md): Instantly clone namespaces with copy-on-write branching - [Sharding](https://turbopuffer.com/docs/sharding.md): Scale a single namespace beyond one index by partitioning across shards - [Embedding](https://turbopuffer.com/docs/embedding.md): Embed documents and queries - [Pinning](https://turbopuffer.com/docs/pinning.md): Reserved compute and disk for predictable cost and performance at sustained high QPS ## Production - [Security](https://turbopuffer.com/docs/security.md): Compliance (SOC 2, HIPAA), encryption, and deployment options - [Encryption](https://turbopuffer.com/docs/encryption.md) - [Private Networking](https://turbopuffer.com/docs/private-networking.md): Configure private con