Free during early access
Local AI, sized to your Mac.
l3ms reads your machine's chip, memory, and GPU headroom, then recommends the models that will actually run well on it. Download, chat, done — everything stays on your Mac.
No cloud. No account. macOS 13 or later.
The actual app — hardware assessment, model library, and a fully local chat.
From “which model?” to a running chat in minutes.
- 01
It checks your hardware
l3ms reads your Mac's chip, memory, and GPU headroom — no setup, no benchmarks to run.
- 02
It recommends a model that fits
Not the biggest model. The one that runs well on your machine, with an honest speed and quality tradeoff.
- 03
You download and run
One click. Inference runs locally through ollama, streaming responses with no uploads.
An app that respects your hardware — and your privacy.
It knows exactly what your Mac can handle.
An 8GB M1 and a 64GB M3 Max should not get the same advice. l3ms measures your real VRAM headroom and recommends accordingly — including how much room a model leaves to spare.
A model library that tells you the truth.
Every model is labeled for your machine: what runs fully on the GPU, what will be tight, and what simply won't fit. No trial and error, no 20GB downloads that crawl at one token per second.
Nothing you type ever leaves your computer.
Prompts, documents, and chat history live on your Mac and nowhere else. Once a model is downloaded, l3ms works entirely offline — the cloud in our icon is crossed out on purpose.
Frequently asked questions
What is l3ms?
l3ms is a macOS app that manages local LLMs. It looks at your Mac's chip, memory, and GPU headroom, recommends models that will actually run well on that hardware, and handles download and inference for you.
Does l3ms send my data anywhere?
No. Prompts and chat history stay on your Mac. Inference runs locally through ollama, and there is no account and no cloud component.
Which Macs does l3ms support?
macOS 13 or later. Recommendations adapt to your actual hardware, so an 8GB M1 gets honest suggestions, not the same list as a 64GB M3 Max.
How is l3ms different from ChatGPT or other cloud AI?
Cloud AI sends your prompts to a server and stores them. l3ms runs models on your own hardware, works fully offline once a model is downloaded, and needs no account.
Does l3ms cost anything?
It is free during early access. Donations are welcome. The app is part of a larger suite of local, offline-friendly apps from AIVI AI Services, an AI consulting firm in Amsterdam.
Do I need technical knowledge to use it?
No. l3ms tells you which model fits your Mac and handles the download. The chat interface is a regular app window, no terminal needed.
Ready when you are.
Free during early access. One download, and your Mac becomes a private AI machine.