Install beancount with Homebrew, apk, apt, dnf, Nix, pacman

Project links

• Homepage
• Repository
• Upstream docs
• Package manager page

Commands and aliases

• bean-check
• bean-doctor
• bean-example
• bean-format
• treeify

Project history

The official history says John Wiegley's Ledger inspired the first Beancount. Blais first experimented with a Python interface to Ledger, then rewrote the idea in Python so one parsed transaction database could serve multiple reports and a web UI.

Beancount version 1 was used continuously by its author for personal, company, and joint-property finances from 2008 to 2012. In 2013, Blais redesigned the system around better inventory and capital-gains handling, a lex/yacc-parsed input language, and a cleaner internal model; this became Beancount version 2.

The README describes three generations: v1 as the original Ledger-like implementation, v2 as the 2013 rewrite and long-running stable branch, and v3 as the stable line since June 2024, trimmed down so several former v2 tools live as independent GitHub projects.

Adoption history

Beancount grew out of real personal-accounting use rather than a vendor product. Its README notes use by many users, companies, and foundations since the project began around 2007, while the history page emphasizes several years of the author's own ledger data as design pressure.

The package is available through many Unix package systems in the supplied metadata, including Homebrew, Debian, Ubuntu, Fedora, Alpine, Nix, and related distributions, which is typical for command-line accounting tools that users want installed beside editors, shell scripts, and version control.

How it is used

Common command-line entry points include bean-check for validation, bean-format for formatting, bean-doctor for diagnostics, bean-example for sample data, and treeify. Users normally edit a ledger file directly, then run Beancount tools against that file.

The Beancount query language adds a SQL-like shell over parsed accounting entries. Its documentation explains that ordinary SQL export was not enough for inventory and lot-aware accounting, so Beancount provides accounting-specific query semantics.

Why package nerds care

Beancount matters to package nerds because it is finance software that behaves like source code: text files, reproducible command-line checks, scriptable reporting, and data that can be kept in Git for decades.

Its history captures a classic Unix-package story: a domain expert liked an existing text tool, rebuilt the model with stricter semantics, and then let the ecosystem split into small installable tools as the core matured.

Timeline

• 2008: Official history records Beancount as begun, after Ledger inspired the first version.
• 2008-2012: The author used Beancount v1 continuously for personal, company, and joint-property finances.
• 2013: Beancount v2 was rewritten from scratch with a new grammar and cleaner inventory model.
• 2014: The official history and query-language documentation were published as part of a larger documentation effort.
• 2020-2024: The README describes v2 as the previous stable version in maintenance mode.
• 2024: Beancount v3 became the current stable version in June, with several former tools moved into independent projects.

Related projects

• Ledger and hledger are the closest related command-line accounting projects; the Beancount docs explicitly compare them and credit Ledger as the original inspiration.
• beanquery is the SQL-like query tool associated with Beancount ledgers. LedgerHub is described in Beancount documentation as an import-automation design related to Beancount workflows.

Sources

• https://beancount.github.io/docs/beancount_history_and_credits/
• https://beancount.github.io/docs/beancount_query_language/
• https://github.com/beancount/beancount#readme
• input.json source_facts.package-manager