Splitpro

Open-source expense splitting, honestly reviewed. No marketing fluff, just what you get when you self-host it.

TL;DR

• What it is: MIT-licensed, self-hosted expense splitting app — a direct Splitwise replacement that puts your financial data on your own server [README][1].
• Who it’s for: Friend groups, roommates, and travelers who hit Splitwise’s free-tier wall (3 expenses per day, mandatory ads) and want a clean, open-source escape hatch [1].
• Cost reality: Splitwise Premium runs ~$20.99/year — cheap in absolute terms. The real case for self-hosting isn’t cost, it’s control: no daily limits, no ads, no risk of another pricing change, and your expense history belongs to you [1].
• Key strength: Clean feature set covering the core Splitwise use cases — multiple split methods, multi-currency, PWA, push notifications, and a Splitwise importer for migration [README].
• Key weakness: This is a small project (1,130 GitHub stars) with gaps that matter: Splitwise import doesn’t pull historical expense amounts (only friends and groups), there’s no debt simplification, the mobile app is PWA-only, and at least one self-hoster tried it and moved to a different tool instead [1][2][README].

---

What is Splitpro

Splitpro is a Next.js web app for tracking shared expenses. You create groups, add expenses, and the app calculates who owes whom. The GitHub description calls it “an open source alternative to Splitwise,” which is exactly what it is — nothing more, nothing less [README].

The project is built and maintained by KM Koushik, a solo developer. It sits at 1,130 GitHub stars as of this review, which puts it firmly in the “promising but niche” tier of self-hosted tools — not a ghost project, but not a community with hundreds of contributors either [merged profile].

The timing of its rise tracks directly with Splitwise’s free-tier crippling. In mid-2024, Splitwise moved to a model where free users are capped at 3 expenses per day and must watch ads. For anyone casually tracking a weekend trip or a month of shared rent, that limit is workable. For anyone running an active house with grocery runs, restaurant splits, and utility bills, it’s a constant irritant [1]. Splitpro exists to be the answer when you search “Splitwise alternative that doesn’t suck.”

One important note upfront: the old public instance at splitpro.app is no longer maintained. If you want Splitpro, you self-host it. There’s no free cloud tier to try first [README].

---

Why people choose it

The comparison landscape here is well-documented by a 2024 feature matrix [1] that benchmarks Splitpro against eight alternatives including Splitwise, Splid, Tricount, Sesterce, and Spliit. That table is the most useful third-party data available on Splitpro, so it’s worth walking through the key contrasts.

Versus Splitwise (the main comparison): Splitwise has every feature you could want — recurring expenses, debt simplification, charts, search, and a rating of 3.5 on the App Store and 4.0 on Play Store from tens of millions of users. The problem is the business model: ads or $20.99/year, and a 3-expense-per-day limit on the free tier that makes casual use painful [1]. Splitpro covers the core use cases — groups, multiple split methods, currencies, receipt uploads, categories — without the paywall or the limit. The feature gaps are real (no simplification of debt, no charts, no search as of the 2024 comparison [1]) but most of them won’t matter for a household or travel group.

Versus Spliit: Spliit is the other major self-hosted Splitwise alternative. Both are MIT-licensed Next.js apps with Docker deployment. The codeberg self-hoster documented in source [2] explicitly archived their SplitPro work-in-progress and switched to Spliit instead — which is a real signal. Spliit gets a mention in the same comparison table [1] and matches Splitpro on most core features while adding debt simplification (which Splitpro lacks) [1]. If you’re evaluating both, Spliit is the alternative you should test before committing.

Versus Tricount / Sesterce: These are polished SaaS apps with strong mobile ratings (4.8/5 across the board [1]), but they’re fully closed-source. Tricount was acquired by bunq; Sesterce runs on a subscription. Neither gives you data portability or the ability to modify the codebase. If you don’t care about self-hosting and just want a better free Splitwise, Tricount or Settleup might be cleaner options with native apps and established communities [1].

On privacy: This is where the pitch lands hardest. Splitwise is a US company. Tricount is Belgian. When you connect expense tracking to a third-party service, your spending patterns, who you live with, and where your money goes are on someone else’s servers. Splitpro self-hosted means that data stays in your infrastructure, subject to your backup policy and your privacy decisions — not Splitwise’s or any acquirer’s [1].

---

Features

Based on the README and the 2024 comparison table:

Core expense management:

• Add expenses with a friend or a group [README]
• Split methods: equal, percentage, share, exact, adjustments, and settlements — more granular than many alternatives [README]
• Categories, currencies, dates, and receipt attachments stored locally [README]
• Negative expenses for refunds and corrections [README]
• Detailed per-person and per-group balance views [README]
• Activity feed showing edits and deletions [README]

Groups and friends:

• Multiple groups or direct friend balances — everything consolidated [website]
• Invite friends by email or from existing friends list [README]
• Group debt simplification is optional (though the 2024 comparison marks it as not supported [1] — check current release)

Mobile and notifications:

• PWA — installable from browser, works like a native app [README][website]
• Push notifications for new expenses and updates [README][website]
• No native iOS or Android app; store ratings are N/A [1]

Multi-currency:

• Balances displayable in a single currency with conversion [README]
• Multiple currency providers with configurable rate limits [README]

Recurring transactions:

• Supported, but requires PostgreSQL with the pg_cron extension — a meaningful setup constraint [README]
• The project publishes a pre-built Postgres image with pg_cron bundled, but if you bring your own database you need to enable the extension manually [README]

Bank integration:

• Experimental feature: load transactions from Plaid and convert them to expenses [README]
• Maintained by a community contributor, not the core developer [README]

Migration and import:

• Splitwise import for friends and groups — partial [README]
• Expenses themselves are not imported, only the people and group structure [README] — a significant limitation if you have years of Splitwise history

Authentication:

• NextAuth only: email magic link, Google OAuth, or OIDC (Authentik, Keycloak, custom) [README]
• No username/password login
• Instance lockdown possible by disabling signups and invites [README]

---

Pricing: SaaS vs self-hosted math

Splitwise (the incumbent):

• Free: 3 expenses per day, mandatory ads [1]
• Premium: ~$20.99/year (as of 2024, EUR pricing) [1]
• 10 million Play Store users, established infrastructure [1]

Splitpro (self-hosted):

• Software: $0 (MIT license) [README]
• Hosting: $5–10/month on Hetzner, Contabo, or DigitalOcean
• Annual hosting cost: $60–120/year

The honest math: Splitwise Premium at $20.99/year is genuinely cheap. On pure cost, self-hosting Splitpro on a VPS is almost certainly more expensive than just paying for Splitwise. If cost is your only concern, Splitwise Premium is probably the right answer.

The case for self-hosting is different: control, privacy, and vendor independence. Splitwise has already changed its free tier once — the 3-expense limit wasn’t always there [1]. Self-hosting means that change can’t happen to you again. Your expense history is under your backup policy, not theirs. And if Splitwise gets acquired (like Tricount was by bunq [1]), your data doesn’t move with it.

Splitpro Cloud: The old public splitpro.app instance is no longer maintained [README]. There is no managed cloud option. Self-host or don’t use it.

---

Deployment reality check

The deployment story is Docker Compose with a PostgreSQL database. The README is clear about the setup path: copy .env.example, configure auth, database, and uploads, start the stack [README].

What you actually need:

• A Linux VPS with at least 1–2GB RAM
• Docker and docker-compose
• A domain and reverse proxy (Caddy or nginx) for HTTPS
• PostgreSQL — bundled in the default compose file or external
• An OAuth provider or SMTP for email magic links (username/password is not supported) [README]
• If you want recurring transactions: PostgreSQL with pg_cron — the project ships a bundled image [README]
• If you want local file uploads to persist: a properly mounted volume for receipts [README]

What can go sideways:

The pg_cron requirement for recurring transactions is the most likely friction point. It doesn’t support cron ranges or lists [README], and if you bring your own managed PostgreSQL (e.g., from a cloud provider), enabling the extension may require a support request or may not be available at all.

Authentication is another surface: there’s no fallback to username/password. If your SMTP goes down or your OAuth app credentials expire, users can’t log in. For a household tool this is manageable, but it’s worth having the OIDC or Google OAuth set up reliably before sharing the instance with non-technical family members.

The codeberg self-hoster in source [2] marked their Splitpro setup as “work-in-progress new Docker deployment” under the archived section — meaning they tried it and eventually moved to Spliit instead [2]. That’s one data point, not a trend, but it suggests the setup has enough friction that some self-hosters evaluate and switch.

Realistic time estimate: 45–90 minutes for a technical user familiar with Docker Compose and reverse proxies. For a non-technical founder following a guide: 3–5 hours including domain, SMTP, and OAuth setup.

---

Pros and cons

Pros

• MIT license. Fork it, modify it, embed it — no lawyer required [README].
• Covers the core Splitwise use cases. Groups, multiple split methods, multi-currency, receipt uploads, push notifications — the daily-driver features are there [README][website].
• Multiple split methods. Equal, percentage, share, exact, adjustments — more granular than simpler alternatives [README].
• Splitwise migration path. Friends and groups can be imported, so you don’t start from zero [README][website].
• PWA with push notifications. Works acceptably as a mobile app experience without being in any app store [README].
• Bank transaction integration. Plaid import to convert transactions into expenses is a genuinely useful feature not common in self-hosted alternatives [README].
• Currency conversions. Multi-currency expenses and group balance normalization built in [README].
• Docker Compose deployment. Standard infrastructure, no exotic dependencies [README].
• Data sovereignty. Your expense data on your server. If Splitwise changes its model again, it doesn’t affect you [1].

Cons

• No historical Splitwise expense import. Friends and groups migrate; expense amounts don’t [README]. If you have years of history in Splitwise, you’re starting clean.
• No debt simplification (as of the 2024 comparison [1] — verify against current release). Splitwise has this; Spliit has it; Splitpro didn’t at time of writing.
• PWA only, no native app. Play Store and App Store ratings are N/A [1]. PWA is decent but not identical to a native app experience, especially for push notifications on iOS.
• No search, no charts (per 2024 comparison [1]). For a household tracker these are convenience features; for any kind of financial visibility they’re meaningful gaps.
• Recurring transactions require pg_cron. A non-trivial database configuration requirement, and the feature has known limitations [README].
• Authentication requires an external provider. No username/password. SMTP or OAuth must be working for anyone to log in [README].
• Solo developer project. 1,130 stars, one maintainer [merged profile]. Bank integration is maintained by a community contributor, not the core team [README]. Bus factor is real.
• No managed cloud option. The public instance at splitpro.app is offline [README]. Self-host or nothing.
• Spliit is a credible alternative that covers similar use cases with debt simplification, and at least one documented self-hoster tried Splitpro and moved to Spliit instead [2].

---

Who should use this / who shouldn’t

Use Splitpro if:

• You’ve hit Splitwise’s 3-expense-per-day limit and don’t want to pay for Premium.
• You want your household or group expense data on your own infrastructure, not a US company’s servers.
• You’re comfortable with Docker Compose and can configure OAuth or SMTP once.
• The missing features (search, charts, debt simplification) don’t matter for your use case — and for most shared-expense tracking they won’t.
• You want MIT-licensed software you can modify or extend.

Consider Spliit instead if:

• Debt simplification is important to your group (fewer individual payments).
• You’ve looked at both and Spliit’s features fit better.
• You want a slightly more established self-hosted option with community momentum.

Stay on Splitwise Premium if:

• You have years of Splitwise expense history you need access to.
• You want a polished native mobile app with high store ratings.
• The $20.99/year is genuinely easier than maintaining a VPS.
• Your group includes non-technical members who need reliable login without any OAuth configuration effort.

Pick Tricount or Settleup if:

• You don’t care about self-hosting and want a polished closed-source alternative to Splitwise with better free-tier terms.
• Native app quality and App Store/Play Store ratings matter to your group [1].

---

Alternatives worth considering

• Spliit — the closest self-hosted alternative. Also MIT-licensed, Next.js, Docker-deployable. Adds debt simplification. Worth comparing directly before choosing [1][2].
• Splitwise — the incumbent. Premium at $20.99/year is cheap. Larger feature set, 10M+ users, native apps. Closed source, US company, free tier now crippled [1].
• Tricount — SaaS, acquired by bunq, 4.8/5 ratings on both stores, 10M+ Play Store users. No self-hosting [1].
• Sesterce — French SaaS, GDPR-compliant, strong ratings, subscription model [1].
• Settle Up — SaaS with ads and subscription, strong mobile ratings, Splitwise import available [1].
• Firefly III — if your real problem is personal finance tracking rather than expense splitting, Firefly III is the self-hosted standard. Different use case, much deeper feature set.

---

Bottom line

Splitpro is the right tool for a specific, narrow problem: you need a Splitwise replacement that runs on your infrastructure, doesn’t limit your daily expenses, doesn’t serve you ads, and won’t change its model on you next year. For that use case, it works. The core expense-tracking feature set is solid, the Docker deployment is standard, and the MIT license means you own it.

The honest caveats are worth repeating: historical expense import from Splitwise doesn’t work, the mobile experience is PWA-only, debt simplification is absent, and this is a solo-developer project — not a company with a support team. Spliit is a credible alternative at the same technical level, and at least one self-hoster evaluated both and chose Spliit [2]. The choice between them mostly comes down to whether you need debt simplification and which codebase you prefer to maintain.

If you’re tired of Splitwise’s increasingly unfriendly free tier and comfortable running a Docker Compose stack, Splitpro is a reasonable afternoon project. If you want someone else to deploy and maintain it, that’s exactly the kind of one-time setup that upready.dev handles for founders who’d rather not spend the afternoon on it.

---

Sources

1. Finding a Splitwise Alternative — desentropia.com (2024-08-29). https://desentropia.com/2024/08/29/finding-a-splitwise-alternative/
2. pleonex/self-hosting: Document and scripts for my own self-hosted services — codeberg.org. https://codeberg.org/pleonex/self-hosting

Primary sources:

• GitHub repository and README: https://github.com/oss-apps/split-pro (1,130 stars, MIT license)
• Official website: https://splitpro.app