A Simple Web Interface for Demonstrating Local LLMs

I built a small tool that makes it easy to share a locally running LLM with other people.

The setup is handled through a simple GUI: select your Ollama model, set a password and instructions, install Cloudflared if needed, and start the server. The app then creates a public access link that you can immediately share with users.

It also includes a request queue, so several people can access a model running on a single machine without sending inference requests simultaneously.

The main goal is to make local LLM demonstrations as simple as possible: configure the demo, click start, and share the link.

Later, I will add running modes for transformer scripts and local Ollama models.

You can find the app and installation guides here.