fromleda/text-humanizer

text-humanizer is an open-source project designed to convert text generated by AI to its "humanized" version. Bypasses the most of AI detectors such as Turnitin or GPTZero

What it solves

It addresses the issue of AI-generated text being easily detectable by AI detectors. It transforms AI-written content into more natural, human-like text while preserving the original meaning across multiple languages.

How it works

The project uses a multi-step rewriting pipeline to introduce linguistic variation:

  1. LLM Rewrite: DeepSeek LLM rewrites the text and translates it into Chinese as an intermediate step to break original patterns.
  2. Google Translation: The text is translated into Turkish to introduce syntactic distortion.
  3. DeepL Translation (Optional): If configured, the text is translated into Japanese for further diversity.
  4. Final Reconstruction: DeepSeek LLM translates the text back into the original language, cleaning up artifacts and restoring natural flow.

Who it’s for

Users who want to make AI-generated content less detectable and more natural in style and tone across eight supported languages.

Highlights

  • Bypasses most AI detectors.
  • Supports 8 languages (English, Japanese, Chinese, Korean, German, French, Spanish).
  • Fully free and open-source.
  • Uses a combination of LLMs and translation engines (DeepSeek, Google Translate, DeepL).

Related

  • Project
    lynote-ai/humanize-text

    A Python toolkit that uses a multi-step LLM and translation chain to rewrite AI-generated text into more natural, human-like drafts.

  • Dispatch
    Detecting LLM-Generated Text with Classical Machine Learning – Results and Insights

    A scikit-learn TF‑IDF + LinearSVC pipeline can distinguish modern LLM‑generated Chinese text from human‑written text with ~85% sentence‑level accuracy, demonstrating that “classical” ML still works for AI‑text detection.

  • Project
    larashero3-dotcom/lieflat-less-ai-tone

    A linguistic research project and toolset that quantifies "AI flavor" in Chinese text and provides a set of rules to remove these markers for more human-like writing.

  • Project
    KKKKhazix/human-writing

    A framework and toolset for AI agents to eliminate "AI-style" Chinese writing by enforcing human-like rhythms, factual substance, and structural checks.

  • Project
    Aspirin0000/zhouli-translator

    A Chinese meme generator and translator that uses DeepSeek to convert modern text into a formal, pseudo-ancient "Zhouli" style and decode it back into plain language.