Open Source Rocket Money Alternatives

Maybe is a personal finance app that runs on your own server instead of a vendor's. The company behind it pivoted and shipped a final working release, so the repository is archived and no longer maintained, but the code stays available under AGPLv3 for anyone who wants to run or fork it.

Self hosting goes through Docker Compose, using a sample compose file, a generated secret key and a PostgreSQL database. Developers can instead run it locally with Ruby and Postgres, then load demo data through a rake task.

That final release documents what the self hosted app covers.

• Net worth dashboard: a trend over time plus your current personal balance sheet.
• Transactions: filtering, search, bulk updates, categories, tags and merchants.
• Budgets: monthly views with category averages and an income summary.
• AI chat: an assistant that knows your finances and answers questions about them.
• Imports: CSV for accounts, transactions and trades, with multi currency support.

Households are handled directly. You can invite members, protect the account with two factor authentication, and build automations against the API using managed keys. Rules auto categorise transactions and detect merchants as data arrives.

Anyone forking the code has to keep the AGPLv3 license and state clearly that the fork is not affiliated with or endorsed by Maybe Finance. The Maybe name and logo are trademarks, so a fork needs its own branding before it goes anywhere public.

Sure is a personal finance app kept alive by the people who use it. The Maybe Finance team spent 2021 and 2022 building a full-featured finance and wealth management product, stopped development when the business did not work out, open-sourced the code, briefly ran a hosted version, then wound that down too. Sure is the community fork that picks it up from there.

It is a working application rather than a starting point, and it needs Ruby at the version pinned in the repository, PostgreSQL above 9.3 and Redis above 5.4.

A few things matter before adopting it.

• Self-hosting: Docker is the documented path, with one-click options on PikaPods, Railway and Hostim.
• Client options: a browser, a macOS desktop app, a mobile app, API clients and LLM agents all reach the same instance.
• Performance transparency: a public dashboard publishes slow requests from the demo site with stack traces attached.
• Naming rules: the Maybe trademark and logo cannot be reused in forks, so a fork needs its own identity.

Local development starts by copying the example environment file and running the setup and dev scripts, with optional demo data to load. Setup guides cover macOS, Linux, Windows and dev containers.

This suits someone who wants their financial data on hardware they control and is comfortable running a Ruby app with Postgres and Redis behind it. Performance on data-heavy views is an open area where contributions are welcome, and the code is AGPLv3.

Financial Freedom is a personal finance application you run yourself, put forward as an alternative to Mint and YNAB. The thing it removes is the intermediary that has to see every transaction in order to categorize it, along with the incentive that intermediary has to sell what it learns.

The application is built on Laravel with MariaDB for storage, Redis, Meilisearch for search and Tailwind CSS for the interface, and it is packaged for Docker. That means it runs on anything that runs containers, from a cloud instance to a Raspberry Pi in a basement, and the data sits on disks you own.

The feature set is aimed at everyday budgeting rather than portfolio management.

• Any bank: import from any institution that can export a CSV file.
• Tagging and splitting: label transactions and divide a single one across several categories.
• Cash flow and budgets: track money in and out against the limits you set.
• Global currencies: work in something other than dollars.
• Automated sync: planned through services such as Plaid or Finicity, and not shipped yet.

Anyone planning to deploy this should know that the project is on hold. The two person team behind it paused development to fund the work through other products, and asks people not to install it until that notice is lifted, because a half finished refactor currently renders a blank screen. Reading and forking the code is still encouraged, and the repository remains the place where progress, discussion and sponsorship happen.

Most budgets fall apart on the bills that land between paychecks. Monetr works the other way around, setting money aside for each recurring expense as its due date approaches, so the number on screen is what is genuinely free to spend before the next paycheck arrives.

It takes its cues from Simple, the banking app that shut down, and rebuilds that idea as software anyone can run. Account data arrives through Plaid for automatic updates, or as OFX files exported from a bank by hand for people who would rather not link anything.

The application stays close to a few ideas rather than growing into an accounting suite.

• Recurring expenses: each one is funded gradually on whatever interval it repeats.
• Paycheck budgeting: allocations follow your pay cycle instead of a calendar month.
• Free to spend: the headline balance is what remains once every commitment is covered.
• Forecasting: projected balances show when a shortfall is coming and when it clears.
• Mobile first: the interface installs as a web app on both desktop and phone.

Self-hosting is free and documented, and the whole source tree is public. Contributors need their own Plaid sandbox credentials, after which a single make develop command brings up the local environment and make clean tears it back down. The hosted version starts with a thirty day trial.

Recurring charges hide in email. Suprascribe reads a connected mailbox, pulls out the receipts and renewal notices, and turns them into a subscription dashboard, so the trial you forgot and the annual plan renewing next week sit in the same list with their real cost.

It is a Next.js 15 application in TypeScript, with Supabase Postgres and row-level security keeping user data separate, Supabase Auth covering Google, Microsoft and Apple sign-in, and the Vercel AI SDK abstracting more than ten model providers. One service fetches mail through the Gmail API, Microsoft Graph or IMAP, and another asks a model to extract subscription details from the message text.

The tiers split along what discovery costs to run.

• Free features: the dashboard, unlimited manual entries, multi-currency support and bring your own key discovery.
• Paid features: automatic discovery from Gmail, Outlook, iCloud or any IMAP account, full history, search, grouping, quick unsubscribe, a calendar view and renewal reminders.
• Key storage: model keys you supply are encrypted at rest with AES-256-GCM and decrypted only while a request runs.
• Mail safety: outbound requests to mail servers refuse private and loopback ranges along with cloud metadata endpoints.
• Payment safety: Stripe webhooks are checked against their signature, and an idempotency guard stops a duplicate tier upgrade.

Running your own copy needs Node.js 20, Yarn v4 and a Supabase project, with Stripe, Resend and Brandfetch optional. The AGPL-3.0 license means a modified copy offered as a network service has to publish its source to the people using it.