Skip to content

Private AI you can run on your own computer

These tools run AI models on your own machine, so prompts need not leave it. They differ in licence, interface and hardware needs, so the notes state each one.

  1. 1

    Open-source desktop app for running AI models locally, with optional connections to cloud model providers.

    Free chat app under Apache-2.0. Needs macOS 13.6+, or Windows 10+ with 8GB RAM; the docs list 6GB VRAM as the GPU minimum. Can also connect to cloud providers with your own keys.

    • Free
    • Open source
  2. 2

    Desktop app for downloading, chatting with and serving open AI models on your own computer.

    Free for home and work use, but the app is closed source; only the CLI and SDKs are MIT. Docs recommend 16GB RAM; macOS 14+ on Apple silicon, Windows, or Linux AppImage.

    • Freemium
  3. 3

    Command-line app and local server for downloading and running open AI models on your own computer.

    MIT-licensed model runner and local server, mainly command-line. Needs macOS 14+ or Windows 10 22H2+. Local use is unlimited; cloud models are an optional paid extra.

    • Freemium
    • Open source
  4. 4

    Self-hosted chat interface that connects to Ollama and OpenAI-compatible model servers.

    Browser interface for Ollama and OpenAI-compatible APIs. Licence is not OSI-approved: branding must stay unless you have 50 or fewer users, permission or an enterprise licence.

    • Free
  5. 5

    Open-source node-based interface and engine for running generative AI models on your own machine.

    Node-based image and video tool, GPL-3.0. Runs on Nvidia, AMD, Intel or Apple Silicon, or CPU-only (slower). Manual install needs Python 3.12 or 3.13; RAM and VRAM minimums are not stated.

    • Freemium
    • Open source

One new comparison, every week

Two tools side by side with our verdict, in one short email. Unsubscribe any time.

Last updated 16 October 2026. How we make money