CivitAIArchive

    About

    CivitAIArchives.com was created in response to the sudden removal of AI models from CivitAI. Our mission is to preserve access to these valuable resources and ensure that the community's work remains accessible for creators and users alike.

    Current Status

    CivitAIArchives.com is under active development. We're working quickly to implement crucial features following the recent CivitAI changes. This page outlines both currently available features and our development roadmap.

    How It Works

    CivitAIArchives.com uses SHA256 file hashes as the foundation of our archival system. This approach allows us to:

    • Verify file authenticity across different hosting platforms
    • Match files from various sources even when filenames differ
    • Create a drop-in replacement for applications that previously relied on CivitAI

    Features: Currently Available

    • SHA256-based file matching to identify and connect models across platforms
    • Basic path compatibility with original CivitAI URLs for known files

    Features: Under Development

    File Discovery & Access (Coming Soon)

    • Direct path compatibility with all original CivitAI URLs
    • SHA256 lookup endpoint at CivitAIArchives.com/sha256/XXX showing all available mirrors
    • Multiple source types including direct downloads, Huggingface repositories, and torrents

    Community Contributions (In Development)

    • Huggingface integration allowing users to add their repository IDs for automatic crawling
    • Metadata recovery system for preserving crucial model information
    • Verification system for confirming file and metadata accuracy

    Resilience & Availability (Planned)

    • Multiple mirror support to ensure files remain available even if individual hosts remove content
    • Torrent integration for distributed file storage that's resistant to takedowns
    • Health metrics showing availability status across different sources

    How You Can Help

    Once our contribution systems are in place, you'll be able to:

    1. Add your Huggingface repositories containing preserved models
    2. Submit metadata for models you've used and want to preserve (Coming Soon)
    3. Provide alternative sources for models you've archived (Coming Soon)
    4. Verify model integrity by confirming SHA256 matches and metadata accuracy (Coming Soon)

    Contact

    For questions about our development progress or to follow updates, please check our discord channel.

    CivitAIArchives.com is not affiliated with CivitAI. We are an independent community initiative created to ensure continued access to community-created AI resources.