Anthill Spider

About

NEW:

-latest update 1.3.1
-Complete refactor of codebase. Anthill Spider now extracts all conversations or a selected conversation, or a selected N conversations.
-Additionally, created Anthill-Collector, for scraping images (including uploads, backgrounds, etc.) from ChatGPT history (search extensions for Anthill Collector ) .


OLD:
______________
 1.0.1
    -added auto-scrolling of entire conversation history to load all conversations into DOM, then full history download (it might be slow but it works!)


Anthill Spider is a privacy-focused tool that helps ChatGPT users export their conversation history for personal use. Create custom datasets for AI training, archive valuable conversations, or analyze your ChatGPT usage—all while keeping your data entirely on your own device.

Features:
🕷️ Scrape all conversations (including lazy-loaded ones).
🔗 Extract prompt/response pairs automatically.
📊 Export to .JSON/.JSONL formats.
🚀 Ready for fine-tuning pipelines (phi-2, Llama, etc.).
🔒 100% local - no data sent to servers.

Perfect for:

    AI researchers building custom datasets.

    Developers creating personalized AI assistants.

    Users wanting to archive their ChatGPT history.

    Anyone interested in AI/ML experimentation.

How it works:

    Install extension (or load unpacked from //chrome:extensions once in DeveloperMode).

    Open ChatGPT in Chrome.

    Click Anthill Spider icon.

    Start export (opens conversations in background tabs).

    Download your data as .JSON.

    Use included pipeline scripts to prepare for training.

Privacy First:

    No data collection.

    No external servers.

    No tracking or analytics.

    Everything stays on your computer.

Source

https://github.com/WintermuteEntertainment/Anthill