OpenModelDB/open-model-database

An open and free database for AI models

What it solves

OpenModelDB is a centralized, community-driven database designed to solve the fragmentation of AI upscaling models. It provides a single location where users can discover, compare, and download image and video upscaling models based on detailed metadata and visual evidence.

How it works

The project maintains a structured database of over 600 AI upscaling models. It uses a React and Next.js frontend to allow users to filter models by architecture, scale, and tags, and view model-specific metadata such as training details and licenses. Users can compare models side-by-side using example images to determine which model best suits their needs.

Who it’s for

Developers, artists, and AI enthusiasts who need to find the most effective AI upscaling models for image and video enhancement.

Highlights

  • Extensive Library: Access to over 600 AI upscaling models.
  • Visual Comparison: Side-by-side comparison of models using example images.
  • Detailed Metadata: Information on training details, architecture, and licenses.
  • Community-Driven: Open-source and community-contributed model entries.

Related

  • Project
    anomalyco/models.dev

    An open-source database of AI model specifications, pricing, and capabilities accessible via a public API.

  • Dispatch
    Models.dev: Building an Open-Source Standard for AI Model Specifications

    An exploration of Models.dev, a community-driven database of AI model pricing, capabilities, and specs designed to provide a unified reference for developers.

  • Project
    wide-trace/open-higgsfield

    An open-source, self-hosted studio for generating images and videos using 32 different AI models via a unified prompt interface and user-provided API keys.

  • Project
    charmbracelet/catwalk

    A community-supported database of AI models and providers compatible with the Crush framework.

  • Project
    deepspeedai/DeepSpeedExamples

    A collection of practical examples, benchmarks, and applications demonstrating how to use the DeepSpeed library for training, inference, and model compression.