DuckDB launches Quack HTTP protocol for multi‑user client/server analytics

DuckDB announced Quack, a new HTTP-based client/server protocol that lets multiple DuckDB instances and applications connect to and share a single database over the network, extending DuckDB beyond its traditional embedded, single — process use. The release preserves DuckDB’s native on-disk format and full SQL compatibility, and optimizes small queries by allowing a single network round trip for sending and receiving results; DuckDB says Quack can move large datasets about 3.5× faster than Arrow Flight and was significantly faster than PostgreSQL for the tested workloads.

Quack exposes DuckDB databases over standard HTTP and transfers data using DuckDB’s native format rather than adopting Arrow Flight SQL or its RPC framework. The DuckDB team said that decision gives them full control over data transfer semantics and protocol evolution. Quack is published under DuckDB’s permissive MIT License and is designed to add remote access and concurrent — user features while keeping DuckDB’s lightweight deployment model.

Developer reaction has been largely positive. On Hacker News participants described Quack as an important step toward shared analytics that retains DuckDB’s ease of deployment; Ryan Glover, managing principal at Lattice Engineering, said the protocol addresses horizontal scaling concerns. A Reddit user, kvlonge, wrote that running DuckDB on a server and allowing remote connections "will be such a big unlock."

The project is intended to enable new topologies: DuckDB instances talking to each other, browser tabs or notebooks querying remote DuckDB servers, and DuckDB serving as a remotely accessible catalog when paired with DuckLake. Amir Sefati noted that combining Quack with object storage, Parquet, DuckLake and vector databases offers a pragmatic architecture for modern AI and data engineering workflows. DuckDB plans continued development through 2026, including integrating Quack into DuckLake, improving performance, and delivering a production — ready release with DuckDB 2.0 later in the year. Roadmap items include better support for remote databases, higher transaction throughput, customizable protocol extensions, and replication features aimed at making Quack suitable for production multi — user deployments.