Home / Money & Finance / AI Transaction Categorizer
AI Transaction Categorizer
Label every payment in a bank statement (groceries, transport, bills and more) with a small AI model on your device.
Free, no uploads, no sign-up. Runs 100% in your browser.
How AI Transaction Categorizer works
Bank files and numbers are read and worked out in your browser. Nothing is uploaded, so account numbers and spending stay on your device. The AI mode compares the meaning of each payment text with your category list using MiniLM sentence embeddings (Transformers.js) in a Web Worker. Keyword rules run first for well known names, and Quick mode uses only the rules, with no download.
Input: OFX, QIF, XML, STA, CSV. Output: CSV, XLSX, QIF, JSON.
How to use AI Transaction Categorizer
- Drop a statement.
- Edit the category list if you like.
- Press Label and wait for the model (25 MB, once).
- Fix any label in the table and download.
Tips
- Bank files hold account numbers, so tools that run on your device are the safe choice.
- Always compare totals with the original statement before you file anything.
- OFX and QIF files import into most money apps.
AI Transaction Categorizer FAQ
- Are my files uploaded?
- No. This tool runs entirely in your browser tab. Files are read from your disk into memory and the result is generated locally.
- Is it free? Are there limits or watermarks?
- Free and open source (MIT). There are no file-size caps, daily quotas, accounts or watermarks.
- Does it work offline?
- Yes, once the page (and any engine it needs) has loaded, it keeps working without a connection thanks to the service worker cache.
Related tools
- Statement Analyzer: See money in and out by month, spending by category, top payees and repeating subscriptions from any bank statement.
- Zero-shot Text Classifier: Sort messages, tickets or headlines into labels you choose, with no training, on your device.
- Bank Statement Converter: Convert OFX, QFX, QIF, CAMT.053 XML, MT940 and CSV bank files to CSV, Excel, JSON, OFX or QIF.
- AI Receipt Scanner: Photograph receipts and get the shop, date, tax, total and line items as CSV or JSON, read by on-device OCR.