A private finance app with several ways in
Wealthfolio combines two products that are often separate. It tracks investments and performance across accounts, but it also records spending, net worth, goals, and allocation plans. The application stores data locally in SQLite and does not require an account. That makes it appealing to people who dislike handing a complete financial history to another hosted dashboard.
The easiest experience is the desktop build for Windows, macOS, or Linux. There is also an iOS app and a Docker-based web server. Manual transactions and CSV import remain free. Wealthfolio Connect is an optional subscription for read-only synchronization with more than 30 financial institutions and encrypted device sync.
The feature set is already broad. Wealthfolio tracks multiple accounts and asset types, calculates time-weighted and money-weighted returns, compares a portfolio with benchmarks, handles currencies and exchange rates, and manages investment activity.
Local-first is meaningful, with qualifications
The desktop application keeps holdings, transactions, account details, goals, and contribution limits in a local database. Core application secrets use the operating system keyring, and addons receive a separate secrets API. This design reduces dependence on a vendor account and makes the data file tangible. Users can keep manual tracking completely offline from brokerage connections.
Local storage also transfers responsibility. A laptop failure can take the only database with it unless the user maintains backups. An open iOS issue reports that database backup and export create 0 KB files on iPhone and iPad, although the reporter says macOS exports work. Until that is confirmed fixed, iOS users should not treat an in-app export as a backup without checking its size and restoring it elsewhere.
Connecting a brokerage changes the privacy boundary too. The README describes Connect as read-only and encrypted, but institution synchronization necessarily involves an external service. Anyone choosing Wealthfolio specifically to keep every financial detail on one device should stay with manual entry or CSV imports and evaluate every addon before granting permissions.
The addon model is unusually well developed. A TypeScript SDK provides typed data access, custom pages, navigation, real-time events, secret storage, and hot reload. Addons declare permissions and require consent. Official examples cover goals, investment fees, and swing trading. This gives technical users a supported extension path instead of forcing changes into the core application.
Desktop is simple, Docker is an operations task
Downloading a desktop build is the sensible first trial. Building it requires Node.js, pnpm, Rust, Tauri, environment configuration, and platform build dependencies. That source path is appropriate for contributors, not necessary for an ordinary user.
The official Docker image serves the frontend and an Axum REST API on one port. A durable deployment needs mounted storage for the SQLite database and encrypted secrets, a generated 32-byte secret, authentication, correct CORS configuration, and HTTPS through a reverse proxy. Password authentication uses Argon2id hashes. OpenID Connect supports allowlists by email or subject and refuses to start without an allowlist unless the operator explicitly enables access for any authenticated identity. Those safe defaults are reassuring, but they require careful configuration.
The server remains a single-user-style SQLite application rather than a general financial database service. PostgreSQL support is currently a feature request, as are multiple separate profiles for people managing several portfolios. A family, adviser, or shared organization should verify the access model before exposing one server to several users.
Financial calculations deserve skepticism
Version 3.6.3 is explicitly a performance and correctness release. It replaces memory-heavy daily valuation clones with a sparse timeline, reduces repeated performance calculations, and shrinks databases by removing duplicated lot data from snapshots. The release also fixes holdings-mode deposits counted as gains, missing quote days, stale rebalance totals, blank currencies during CSV import, and time-weighted returns broken by dormant days. These are valuable repairs, and they reveal how hard the accounting model is.
Current issues show that edge cases remain. Issue 1414 reports a portfolio-level time-weighted return becoming implausible when accounts have different starting dates and large contributions. Issue 1425 says a normal negative cash balance before T+2 settlement can make the entire period return null in 3.6.3. Another report says refunds reduce spending correctly but inflate performance. These are detailed reproductions, not proof that every portfolio is wrong, but financial software should be judged by its worst edge cases.
Cost basis is another hard boundary. An open request says FIFO is the only method, while investors in some jurisdictions need weighted-average or LIFO calculations. Wealthfolio can still be a useful personal tracker there, but its figures should not be substituted for broker statements or tax software. Reconcile balances, contributions, realized gains, and returns against an independent source before making decisions.
Imports take work as well. The general CSV mapper is useful, but an open request explains that many brokerage exports need preprocessing, and newer requests cover Fidelity NetBenefits and Trading 212 currency conversion. Before entering years of history, test a small file, review every mapped activity, and retain the original statements.
Health and the decision
The repository was pushed on August 8, 2026, one day after v3.6.3 shipped. It has 373 open issues and pull requests combined, with new reports and development activity continuing through August 10. Releases arrived repeatedly across May, June, July, and August. This is clearly an active project, while the large queue reflects both strong adoption and a wide surface of calculation, platform, import, and feature requests.
Documentation is strongest for developers and self-hosters. The README carefully explains Docker secrets, cookies, CORS, OIDC, volumes, source builds, architecture, and addons. End-user guidance is less complete relative to the number of financial concepts the application exposes.
Wealthfolio earns a recommendation for privacy-minded individuals who want a capable tracker and accept some reconciliation work. Start with the desktop app, import one account, verify its math, and establish a tested backup. Self-host only when browser access is worth the added security and maintenance duties.