Find tools › Open-source projects › Keep data private

project-maya

Open-source project · by mw00

Run a massive 321B AI model locally on your own graphics cards without sending data to the cloud.

Some setupWell known

What it costs

Free and open source; you may pay for the AI model it uses.

Use it to

  • Chat privately with a large local AI model
  • Host a local OpenAI-compatible API for coding agents
  • Monitor GPU and memory usage during local model inference

How to start

  1. Clone the repository using git clone https://github.com/mw00/project-maya.git && cd project-maya

  2. Run ./setup.sh and follow the prompts to install dependencies and download the model

  3. Open http://127.0.0.1:8080 in your browser to use the chat and live monitor

Good to know. Pushed over 4 months ago. Requires heavy hardware (high VRAM, fast NVMe SSD, and lots of RAM) and takes substantial setup time.

Before you install

  • Found automaticallyNot reviewed by a person yet. Check it before installing.
  • License: MITYou are allowed to use it under these terms.
  • Last updated 2026-11-18

More detail

Run the massive 321-billion-parameter GLM-5.3-Flash model privately on your own GPUs.

Runs a datacenter-class model locally on local hardware with zero data leaving your machine, providing an OpenAI- and Anthropic-compatible API for your local tools and coding agents.

Similar tools

Descriptions are written in plain English from each project's own documentation and refreshed by our agent; this one was last checked 2026-10-14. Spotted a mistake? · Open in Find tools · © 2026 Which AI Map