AI Blocks › Build and ship software
Local model converter
Give it a model. Get a smaller file that runs on your laptop or Mac with no GPU, plus the run steps.
Local model converter is a free AI skill for the job "Build and ship software". Give the instructions to an AI assistant that supports the required files and tools, then review its output.
Free · No account · MIT licence · Made by Orchestra Research · Thinking: low
Made by Orchestra Research. See the original on GitHub; we copied it as it is, with its licence.
What it does
A smaller version of an AI model that runs on your laptop, a Mac or a plain CPU with no GPU. Give it a model and get a compact GGUF file plus the steps to run it. It builds llama.cpp for your machine: CPU, Apple Silicon or Nvidia. It converts the model to GGUF, then shrinks it with quantization anywhere from 2 to 8 bits. A smaller number saves space but loses some quality, and an importance matrix helps low-bit files stay accurate. The result works in LM Studio, Ollama and similar tools. It also says when another format fits better, such as AWQ for NVIDIA servers.
You give it
- The model you want to shrink
- Your computer type and memory
- How much quality loss you can accept
- Where you plan to run it
You get
- Build steps for your machine
- A GGUF conversion command
- A shrunk model at the level you pick
- Run steps
- A guide to quality versus size
Use it when
- A model is too big for your laptop memory.
- You use a Mac and want it to run well.
- You need one file that works in Ollama or LM Studio.
- You want to compare different shrink levels.
How to use it
- Add it to your AI once: download it and add the folder to Claude, ChatGPT or Codex as a skill.
- Ask for the job in your own words. Your AI opens the skill when it fits.
- It works best when your AI can run commands and build llama.cpp.