TaxHacker is a self-hosted AI accounting app that automates expense and income tracking for freelancers and small businesses.
The tool addresses the time-consuming work of manual receipt and invoice processing. Users upload photos of receipts, invoices, or PDFs, and TaxHacker uses large language models to automatically extract structured data including product names, amounts, dates, merchants, and taxes. The system stores this information in a database and supports automatic currency conversion using historical exchange rates from transaction dates, including support for cryptocurrencies. Users can define custom fields with their own AI prompts to extract specific information relevant to their accounting needs, and the tool provides filtering, multi-project support, and import/export capabilities.
Adoption suits freelancers, indie developers, and small business owners who handle multiple currencies or need to process high volumes of receipts. The project emphasizes self-hosting and data privacy, allowing users to choose between cloud LLMs like OpenAI, Google Gemini, and Mistral, or run local models through Ollama, LM Studio, vLLM, or LocalAI. The tool handles documents in any language and works with store receipts, restaurant bills, invoices, bank statements, and even handwritten receipts. The README explicitly notes the project is in early development and should be used at your own risk.
Development activity shows ongoing work with regular commits addressing features and fixes. The maintainer is actively seeking employment opportunities and has indicated interest in relocating or remote work, which may affect long-term project maintenance. The project accepts community feedback and appears responsive to reported issues, though the early-stage status means the codebase and features remain subject to significant change.
