Yes, Ledger Chat is 100% free and open source under the MIT license. You bring your own Anthropic or OpenAI API key for the AI chat features. The analysis views work without any API key.

How does this compare to Quicken?

Ledger Chat is not a replacement for Quicken — it's a companion. You continue using Quicken to sync bank transactions and categorize spending. Then export to QIF and use Ledger Chat for the analysis and reporting that Quicken does poorly: multi-year comparisons, recurring charge detection, tag cross-tabulation, AI-powered chat, and interactive visualizations like the Financial Bedrock and Sankey cash flow diagram.

Analyze years of financial data with AI — in your browser

Drop a QIF export, bank CSV, or SQLite database. Get 10 analysis views and AI chat. Your data never leaves your machine.

Explore with demo data Import your own data View on GitHub

Ledger Chat Financial Bedrock view showing spending layers with income comparison and breathing room analysis

Works with exports from: Quicken · GnuCash · Moneydance · Mint · Any bank

Your financial data deserves better reporting

You've spent years categorizing transactions in Quicken. You've built a complete financial history. But when you try to actually analyze it:

Year-over-year comparisons limited to 2 periods
Tag reports broken on split transactions
No recurring charge detection
Budget vs actual doesn't exist on Mac
Dashboard is rigid and useless
Cash flow visualization? What cash flow visualization?

So you export to Excel. Again. And spend an hour building the same pivot table you built last month.

Ledger Chat reads your exported data and gives you the analysis tools Quicken should have built — in seconds, not hours.

10 views Quicken should have built

Dashboard with KPI cards for spending, income, savings rate, and mini treemap

Dashboard — Everything at a glance

Seven KPI widgets with mini visualizations. Click any card to drill into the full view.

Spending grid with categories as rows and months as columns, color-coded deviations

Spending Grid — Compare up to 5 years side by side

Categories as rows, months as columns. Color-coded cells flag deviations from your historical average. Click any cell for the transactions behind it.

Sankey diagram showing income flowing through expense categories

Cash Flow Sankey Diagram — Visualize where your money goes

Income sources on the left, expense categories on the right, flow width proportional to amount. The visualization personal finance enthusiasts share on Reddit.

Year over year comparison across 3-5 years with percentage changes

Year over Year Comparison — Track spending trends across years

Same period compared across 3-5 years. Quicken limits you to 2 periods. Bold highlights flag changes exceeding 25%.

Recurring charges with auto-detected subscriptions and annual cost projections

Recurring Charge Detection — Find every subscription automatically

Auto-detected from your transaction patterns. Annual cost projections and price change alerts. The report that replaces a Rocket Money subscription.

Baseline vs actual spending compared against 12-month averages

Baseline vs Actual — Compare spending to your own history

Your spending compared against your rolling 12-month average. Uses real data instead of aspirational budget numbers.

Net worth stacked area chart showing account balances over time

Net Worth Tracking — See your wealth composition over time

Estimated month-end balances from cumulative transactions. Stacked area chart with account grouping and hover breakdown.

Income vs expenses monthly breakdown with savings rate

Income vs Expenses — Track your savings rate monthly

Monthly breakdown with income, expenses, net, and savings rate. Red-tinted rows when expenses exceed income.

Transaction browser with search, filters, and sort capabilities

Transaction Browser — Search and filter everything

Full transaction table with search, category filters, date ranges, and sort. Click any row to expand split details, memos, and tags.

Financial Bedrock view classifying spending into four layers

Financial Bedrock — See what it costs you to live

Your spending classified into layers: bedrock fixed costs, essential variable, committed obligations, and discretionary. See your true breathing room.

Ask anything about your money

"What did I spend at Amazon this year?" "Compare my grocery spending to last year" "Find all charges that increased in price" "Show me my biggest spending day each month" "Which subscriptions cost the most annually?"

Powered by your own Anthropic or OpenAI API key. The AI queries your data through 12 specialized tools — it doesn't just guess, it calculates.

The 10 analysis views work without any API key. AI chat is optional.

Three steps. Under a minute.

Export your data

QIF from Quicken, CSV from your bank, or any SQLite database with transactions.

Drop the file

Everything is parsed locally in your browser. Your data never touches a server.

Explore and ask

10 views populate instantly. Chat answers any question about your financial history.

Explore with demo data Import your own data

Your data stays on your machine. Period.

No server

No account

No tracking

Ledger Chat is a static website. There is no backend. No database. No user accounts. Your financial data is parsed in your browser and stored in IndexedDB — a local storage system built into every modern browser.

The only external network calls are:

Font files from Google Fonts (IBM Plex)
D3.js and sql.js libraries from CDN
AI API calls to Anthropic or OpenAI using your API key (only the query context is sent, not your full dataset)

Don't trust us? The entire source code is open.

View source on GitHub

Works with the data you already have

QIF	Quicken, GnuCash, Moneydance, KMyMoney, HomeBank
CSV	Any bank export, Mint data, spreadsheet exports, Tiller
QFX/OFX	Bank download files
SQLite	Any SQLite database with financial transaction data

Quicken users: File → Export → QIF. That's it.

Mint users: your old CSV export works as-is.

Bank users: download transactions as CSV.

See what it actually costs you to live

Financial Bedrock landscape showing spending tiers compared against income

Bedrock — Fixed costs that don't change. Mortgage, insurance, car payment, phone.

Essential — Necessary but variable. Groceries, gas, utilities, medical.

Committed — Chosen obligations. Giving, subscriptions, kids' activities.

Regular — Everyday discretionary. Dining, shopping, entertainment.

The gap between your layers and your income is your breathing room — the money available for savings and big purchases.

When a big purchase exceeds your breathing room, the visualization shows exactly how it impacts your savings. It's not daily spending that drains savings — it's 6-8 big purchases per year that each felt reasonable in isolation.

Free forever. Open source. MIT licensed.

Ledger Chat is free to use with your own API key. The source code is public on GitHub. You can read every line, verify the privacy claims, contribute improvements, or fork it for your own use.

Built with vanilla HTML, CSS, and JavaScript. No React. No npm. No build step. Ten files. You could read the entire codebase in an afternoon.

If Ledger Chat helped you understand your finances better:

View on GitHub Buy me a coffee

Frequently asked questions

Is my financial data safe?

Everything runs in your browser. Your financial data is parsed locally and stored in IndexedDB — your browser's built-in local database. No server ever sees your transactions. The only external calls are to the AI provider (Anthropic or OpenAI) using your own API key. Those calls contain only the specific query context — a summary or filtered subset — not your complete transaction history. The source code is open on GitHub. Verify it yourself.

What file formats are supported?

QIF (Quicken Interchange Format), CSV (bank exports, Mint exports, spreadsheets), QFX/OFX (bank download files), and SQLite databases containing financial transaction data. If your finance software can export to any of these formats, Ledger Chat can read it.

Does this replace Quicken?

No. Ledger Chat is a companion to Quicken, not a replacement. You continue using Quicken to sync bank transactions, categorize spending, and manage your accounts. Then export your data as QIF and use Ledger Chat for the analysis views and AI chat that Quicken doesn't provide — multi-year comparisons, recurring charge detection, the Financial Bedrock view, and natural language queries.

Is it really free?

Yes. Ledger Chat is open source under the MIT license. The app itself costs nothing. The AI chat features require an API key from Anthropic or OpenAI, which you manage and pay for directly — typically $0.01-0.10 per query. The 10 analysis views work without any API key at all.

How much does the AI cost to use?

With your own Anthropic API key using Claude Sonnet, a typical query costs about $0.01-0.03. A comprehensive Financial Review using Claude Opus costs about $0.85. You control your spending because you control the API key. Most users spend $1-3 per month.

Can I use it with my old Mint data?

Yes. If you have a CSV export from Mint, drop it directly into Ledger Chat. The parser will detect the column format and parse your transactions. All of Mint's transaction history becomes queryable and visualized.

Does it work with GnuCash, Moneydance, or other software?

Yes. Any software that exports to QIF, CSV, or SQLite is compatible. GnuCash, Moneydance, KMyMoney, HomeBank, and most banking websites provide exports in one of these formats.

What happens to my data when I close the browser?

Your data persists in IndexedDB. When you reopen Ledger Chat, your transactions, chat history, and settings are all restored automatically. You don't need to re-import your file. To permanently delete your data, use the "Clear data" option in settings.

Can I contribute?

Yes. The project is on GitHub. Parser edge cases (different Quicken versions, unusual CSV formats, date format variations) are the biggest need. Visualization ideas, bug reports, and documentation improvements are all welcome.

I found a bug or my file doesn't parse correctly

Open an issue on GitHub with a description of the problem and, if possible, a sanitized sample of the data that caused the issue. Remove real names and amounts — the file structure is what matters for debugging.