TaxHacker — What is it?

TaxHacker is a self-hosted AI-driven accounting tool that automates expense and income tracking by extracting data from receipts and invoices.

⭐ 308 Stars 🍴 51 Forks TypeScript Author: vas3k
Source: README View on GitHub →

Why it matters

TaxHacker is gaining attention due to its AI-powered data extraction from receipts and invoices, addressing the pain points of manual data entry and the need for accurate financial tracking. Its self-hosted nature also caters to privacy concerns, and its support for various AI models and currencies adds to its appeal.

Source: Synthesis of README and project traits

Core Features

AI Data Extraction

TaxHacker uses AI to automatically recognize and extract information from receipts, invoices, and PDFs, including product names, amounts, dates, and merchants, and stores it in a structured database.

Source: README
Multi-Currency Support

The app supports automatic currency conversion for over 170 world currencies and 14 popular cryptocurrencies, using historical exchange rates from the transaction date.

Source: README
Customizable Categories and Fields

Users can create custom categories, projects, and fields to organize transactions according to their specific needs, and write custom AI prompts for more detailed information extraction.

Source: README
Self-Hosted Mode

TaxHacker offers a self-hosted mode, allowing users to keep their financial data on their own infrastructure, ensuring privacy and control over their information.

Source: README

Architecture

The architecture of TaxHacker is inferred to be modular, with a frontend built using Next.js and an API layer. It uses Prisma for database interactions and PostgreSQL as the database. The codebase is structured into various directories for different functionalities, such as AI processing, file handling, and user interface components. The project utilizes Docker for deployment, indicating a containerized and scalable architecture.

Source: Code tree + dependency files

Tech Stack

infra: Docker  |  key_deps: @prisma/client, next, react, react-dom  |  language: TypeScript  |  framework: Next.js, Prisma

Source: Dependency files + code tree

Quick Start

curl -O https://raw.githubusercontent.com/vas3k/TaxHacker/main/docker-compose.yml docker compose up
Source: README Installation/Quick Start

Use Cases

TaxHacker is suitable for freelancers, indie-hackers, and small businesses. It is useful for automating accounting processes, managing expenses and income, and preparing tax reports. Specific scenarios include tracking business expenses, managing multiple projects, and generating tax-ready reports.

Source: README

Strengths & Limitations

Strengths

  • Strength 1: AI-driven data extraction simplifies accounting processes.
  • Strength 2: Self-hosted mode ensures data privacy and control.
  • Strength 3: Customizable features cater to specific business needs.

Limitations

  • Limitation 1: The project is still in early development and may have bugs or limited features.
  • Limitation 2: The effectiveness of AI extraction depends on the quality of the input documents and the AI model used.
Source: Synthesis of README, code structure and dependencies

Latest Release

v0.7.0 (2026-04-03): Added support for local LLMs. v0.6.1 (2026-03-12): Fixed critical issues and improved authentication. v0.6.0 (2025-07-23): Introduced custom LLM support and item detection. v0.5.5 (2025-05-08): Added invoice generator and business details settings. v0.5.1 (2025-04-10): Major bugfixes.

Source: GitHub Releases

Verdict

TaxHacker is a promising open-source project for those seeking an AI-driven accounting solution with a focus on privacy and customization. It is particularly suitable for small businesses and freelancers looking to automate their accounting processes and manage their finances efficiently.

Source: Synthesis
Transparency Notice
This page is auto-generated by AI (a large language model) from the following public materials: GitHub README, code tree, dependency files and release notes. Analyzed at: 2026-04-19 10:11. Quality score: 85/100.

Data sources: README, GitHub API, dependency files