Introducing the AllSquare CLI and API.

Manage AllSquare from your terminal, your own code, or your favorite AI assistant. The new CLI and API launch alongside a much more capable AllSquare GPT and MCP connector.
In April, AI assistants could read your groups and add expenses. Now they can do almost everything you can do in AllSquare, and there are two new ways to connect: a command-line tool and a public API.
The AllSquare CLI (beta)
Manage AllSquare from your terminal, or let coding agents like Claude Code and Codex do it for you. Install it with npm install -g allsquare, then run allsquare auth login to sign in through your browser.
- Work with groups, members, balances, expenses, recurring expenses, and settlements
- Add, edit, delete, and settle expenses, with a confirmation step before anything destructive
- JSON output and clear exit codes built for scripts and agents
- Safe retries: repeat a command with the same idempotency key and nothing is written twice
- Sign-in credentials are stored in your system keychain
The AllSquare API
The new v1 API gives your own apps and scripts the same access. It uses OAuth, supports safe retries with an Idempotency-Key header, and comes with an interactive reference where you can try every endpoint.
Your AI assistant can do a lot more
The AllSquare GPT and MCP connector picked up a long list of new abilities:
- Look up, edit, and delete expenses
- Search your expense history by group, description, payer, or date
- Settle up: preview who pays whom with Venmo links, then mark expenses settled once you confirm
- Edit, pause, resume, and delete recurring expenses
- Create groups and add members by email
- See your balances across every group at once
- Ask spending questions like "What did we spend on groceries last month?" or "Who paid the most?"
- Connect more MCP clients, including ones that sign in from a desktop app or terminal
New docs
The AllSquare docs now cover AI assistants, the CLI, and the API in one place, with copyable setup steps for each. AI tools can read a summary at /llms.txt.
Try it
Start with the CLI guide, the API setup guide, or the AI assistant docs.
Got feedback? We'd love to hear it.


