43 developer tools, always updated — new upstream releases land in apt within hours. Join the Discord

📅 From , the apt mirror will require a subscription. See pricing

🦆 Install Latest DuckDB on Debian

A fast in-process analytical SQL database — SQLite for analytics

Latest version: 1.5.5 · updated 2026-07-22
SourceVersion
deb.griffo.io1.5.5 ✅
Official Debian🚫 not packaged
← Back to home
📌 Installing on a specific release? Debian 12 (bookworm) · Debian 13 (trixie) · Debian 14 (forky) · Debian Sid

What is DuckDB?

DuckDB is a fast, in-process analytical SQL database — think "SQLite for analytics". It runs inside the calling process with no server to set up, queries CSV, Parquet and JSON files directly (SELECT * FROM 'data.parquet'), and speaks a rich, friendly SQL dialect with excellent performance on analytical workloads. This package ships the official duckdb CLI.

⚠️ No server, no daemon: DuckDB is an embedded database — it runs inside your process (or the CLI), so there is no service to configure, no port to open and no systemd unit. Install it and run duckdb. For the same reason, one database file belongs to one process at a time.
🚀 Always ahead of the official archives: Debian and Ubuntu freeze package versions when a release ships — this repository publishes new upstream releases typically within hours, for every supported release including stable. A simple apt upgrade keeps you on the latest version.

⚡ Key Features of DuckDB

📊 Built for Analytics

A columnar, vectorized engine crunches millions of rows per second — aggregations, joins and window functions at OLAP speed.

📁 Query Files Directly

SELECT * FROM 'events.parquet' — CSV, Parquet and JSON are queryable (and writable) without any import step.

🪶 Zero Setup

In-process like SQLite: a single binary, a single optional database file, nothing to administer.

🔌 Runtime Extensions

INSTALL httpfs; pulls signed extensions from the official repository — S3/HTTP access, spatial, full-text search and more.

🗣️ Friendly SQL

SELECT * EXCLUDE, GROUP BY ALL, list/struct types, PIVOT — a modern dialect that removes boilerplate from everyday queries.

🤝 Plays Well with Others

Reads SQLite, PostgreSQL and MySQL through extensions; exports Parquet/CSV/JSON — ideal glue for data pipelines.

🏆 Why install it with apt?

  • Extensions just work: this package uses upstream's official glibc build, so INSTALL httpfs; matches the official linux_amd64/linux_arm64 extension repository
  • Every release, including stable: the official archives carry DuckDB only in testing/unstable — here Bookworm and Trixie get the same current version
  • No curl-to-bin: a GPG-signed repository instead of downloading binaries into /usr/local/bin by hand
  • Automatic updates: new upstream releases arrive with your normal apt upgrade
  • Multi-arch: amd64 and arm64 — the architectures upstream publishes Linux CLI binaries for

📦 Installation from deb.griffo.io

Run the commands

Step 1: Add Repository

sudo install -d -m 0755 /etc/apt/keyrings
curl -fsSL https://deb.griffo.io/EA0F721D231FDD3A0A17B9AC7808B4DD62C41256.asc | sudo gpg --dearmor --yes -o /etc/apt/keyrings/deb.griffo.io.gpg
echo "deb [signed-by=/etc/apt/keyrings/deb.griffo.io.gpg] https://deb.griffo.io/apt $(lsb_release -sc 2>/dev/null) main" | sudo tee /etc/apt/sources.list.d/deb.griffo.io.list > /dev/null
sudo apt update
install -d -m 0755 /etc/apt/keyrings
curl -fsSL https://deb.griffo.io/EA0F721D231FDD3A0A17B9AC7808B4DD62C41256.asc | gpg --dearmor --yes -o /etc/apt/keyrings/deb.griffo.io.gpg
echo "deb [signed-by=/etc/apt/keyrings/deb.griffo.io.gpg] https://deb.griffo.io/apt $(lsb_release -sc 2>/dev/null) main" | tee /etc/apt/sources.list.d/deb.griffo.io.list > /dev/null
apt update

Step 2: Install DuckDB

sudo apt install -y duckdb
apt install -y duckdb

🚀 First run

duckdb                                        # in-memory interactive shell
duckdb mydb.duckdb                            # persistent database file
duckdb -c "SELECT * FROM 'events.parquet' LIMIT 10;"   # query files directly
duckdb -c "INSTALL httpfs; LOAD httpfs;"      # extensions from the official repo

📦 Package Build Repository

The Debian packages are automatically built and maintained in this GitHub repository:

🔗 Related Packages

Also available from deb.griffo.io:

🎯 Perfect for: data engineers exploring Parquet and CSV files, analysts who want SQL without a database server, scripts and pipelines that need fast local analytics, and anyone replacing pandas one-liners with plain SQL.

💝 Support This Project

If this repository saves you time and effort, please consider supporting it!

⭐ Star on GitHub 🐦 Share on Twitter

Just after the command? See apt install duckdb — one page covering Debian and Ubuntu.

📦 Recent releases from this repository

📚 Release-specific install guides

❓ Frequently asked questions

Is DuckDB in the official Debian repositories?

No — DuckDB is not packaged in the official Debian archives. This repository is currently the only apt source, serving DuckDB 1.5.5.

How do I install the latest DuckDB on Debian?

Add the deb.griffo.io repository once using the instructions above, then run: sudo apt install duckdb. New releases arrive through the normal sudo apt upgrade.

Are the packages signed and how are they built?

Every package is signed with the repository's GPG key (EA0F721D231FDD3A0A17B9AC7808B4DD62C41256) and built from upstream releases in public GitHub packaging repositories that anyone can inspect.

Which Debian releases are supported?

Debian 12 Bookworm, Debian 13 Trixie, Forky and Sid.