Lizard LLM

Private local AI for Windows, running on your own PC

Lizard LLM is a hardware-aware local AI runtime for Windows 10 and Windows 11. It checks your processor, memory, and graphics hardware, recommends a GGUF model that fits, and runs the model on your own computer. Once a model is set up, prompts, responses, and model weights stay on the device and inference works without an internet connection.

The Windows application includes two native inference engines, automatic CPU or GPU selection, a local chat interface, and an OpenAI-compatible API for connecting other software. Activation costs $5 once; it is not a subscription.

Download Lizard LLM for Windows or read the complete product and hardware overview.

How Lizard LLM works

  1. Install the Windows application.
  2. Let Lizard inspect the CPU, RAM, and graphics hardware.
  3. Choose one of the compatible local AI models it recommends.
  4. Run private AI chat or connect through the local OpenAI-compatible API.

Built for private, practical local AI

Lizard Native uses Direct3D 12 for resident-GPU inference when a compatible graphics card is available. Caterpillar is designed for CPUs and shared or integrated graphics. Lizard selects an engine for the machine and can fall back to bundled llama.cpp or an optional Ollama installation where needed.

See installation and API documentation, the benchmark methodology, and answers about privacy, pricing, GPUs, and offline use.