Run This LLM
The #1 local LLM hardware calculator on the internet. Pick a model to see the hardware it needs, or enter your hardware to see which models it can run.
The problem
Running an open-weight model locally starts with a question that is surprisingly hard to answer: will it fit? It depends on parameter count, quantization, context length and the GPU's memory, and the answer is scattered across model cards, forum threads and benchmark charts.
Getting it wrong means buying the wrong GPU, or downloading a model that won't load.
What I built
Run This LLM answers it in both directions. Start from a model to see the VRAM it needs, recommended builds and performance estimates. Start from your hardware to see what it can run.
It covers more than 295 models with benchmarks and works from the browser sidebar as a Chrome extension. For people who would rather not build the machine themselves, it takes requests for custom builds specced to the models they want to run.
What it does
Model to hardware
VRAM requirements, recommended builds and performance estimates for each model.
Hardware to models
Enter your hardware and see which models it can run.
Benchmarks
Real performance numbers alongside the estimates.
Chrome extension
Look up any model's requirements from the browser sidebar.
Custom builds
Machines specced, built and shipped for the models a customer wants to run.
By the numbers
Under the hood
- Static site
- Chrome extension
- Installable web app
- Benchmarks
Want a system like this working for your business?
I design, implement, and optimize agentic systems that generate revenue, automate operations, and scale businesses. Tell me what you're working on.









