GankAIGC is an open-source tool designed to address the challenges of AI detection in academic writing. It provides functionalities for "de-AIing" text, enhancing originality, and performing academic polishing, targeting researchers, students, and academics who need to ensure their work appears human-authored and meets academic standards. The primary benefit is improving the perceived authenticity and quality of written content against AI detection tools.

How It Works

The project employs a robust architecture combining FastAPI for the backend, React/Vite for the frontend, and PostgreSQL for database management. It offers several AI processing modes, including de-AIing, polishing, originality enhancement, and emotional tone adjustment, all configurable to use user-provided OpenAI-compatible API endpoints and keys (BYOK). Processing is managed via a PostgreSQL-based task queue, ensuring asynchronous and scalable execution. This approach allows for flexible deployment and user-specific API configurations.

Quick Start & Requirements

GankAIGC supports multiple deployment methods:

• Source Code: Requires Python 3.11+ and PostgreSQL. Installation involves git clone, pip install -r requirements.txt, and running python main.py.
• Docker Compose: A streamlined approach using docker compose up -d after configuring the .env.docker file.
• Windows One-Click: A self-contained package available via releases, requiring extraction to an English-only path before running start.bat.

Official releases can be found at https://github.com/mumu-0922/GankAIGC/releases/latest. The application is accessible via http://localhost:9800, with the admin backend at http://localhost:9800/admin and API documentation at http://localhost:9800/docs.

Highlighted Details

• Multi-Mode AI Processing: Offers distinct modes for de-AIing, academic polishing, originality enhancement, and emotional expression adjustment.
• User Management & Credits: Features an invitation code system for registration and a "beer credit" system for API usage, managed via redemption codes.
• Bring Your Own Key (BYOK): Users can securely configure and use their own OpenAI-compatible API keys and base URLs.
• Project Organization & Export: Supports archiving tasks within projects and exporting results in .docx and .md formats.
• Comprehensive Admin Panel: Includes data dashboards, session monitoring, user management, invitation/redemption code management, announcements, and system configuration.

Maintenance & Community

The README does not detail specific contributors, sponsorships, or community channels like Discord or Slack. It notes that the project is a modification of "BypassAIGC."

Licensing & Compatibility

The project is licensed under CC BY-NC-SA 4.0 (Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International). This license strictly prohibits commercial use without explicit authorization and requires any derivative works to be shared under the same terms, with proper attribution.

Limitations & Caveats

The CC BY-NC-SA 4.0 license imposes a significant restriction, forbidding any commercial application of the software. For Windows users, the one-click installer requires the application to be extracted into a directory path containing only English characters, without spaces or special symbols. Deploying publicly necessitates careful security configuration, including changing default passwords and managing API keys securely. The effectiveness of the AI features is dependent on the quality and configuration of the external AI models used.