ToshLLM field guide

Getting started

Install ToshLLM, choose the correct build, pass the first-launch security check and start a local model.

Updated August 16, 2026 · ToshLLM beta
Before you install

Built for a specific kind of Mac.

ToshLLM is a native macOS application for Intel Macs with AMD graphics. It requires macOS 14 or newer and at least 16 GB of memory. For larger mixture-of-experts models, 32 GB or more gives the model and macOS more room to work.

RequiredIntel Mac, AMD GPU with Metal support, macOS 14 or newer and 16 GB RAM.
Recommended32 GB RAM for larger models and enough free storage for GGUF model files.
Apple silicon is not the target.

ToshLLM exists to bring modern local inference to Intel Macs with AMD GPUs. The application and its bundled engines are built around that hardware.

Step 01

Choose the right download.

The standard release is the right choice for most supported Intel Macs. Older Intel processors without AVX2 require the separate no-AVX2 build. The updater keeps the two variants on separate channels so a no-AVX2 installation does not replace itself with an incompatible build.

  1. Download the appropriate DMG from the ToshLLM release page.
  2. Open the DMG and drag ToshLLM into Applications.
  3. Launch ToshLLM from Applications.

The inference engines are included. You do not need Homebrew, Python or a separate llama.cpp installation.

Step 02

Pass the first-launch check.

Current beta releases may not yet be notarized by Apple. If macOS blocks the first launch, open System Settings → Privacy & Security, locate the ToshLLM notice and choose Open Anyway. Confirm only when the application came from the official ToshLLM release.

This is a first-launch trust decision.

It does not grant ToshLLM unrestricted access to your Mac. Future releases distributed with a stable Developer ID signature and notarization are intended to follow the normal macOS update path.

Step 03

Download and run a model.

Open Models to see recommendations for the detected GPU and available memory. Download a GGUF model, choose Use, then start the server from Home. ToshLLM stores models in ~/models by default; the folder can be changed in Settings.

Begin with a smaller model and its recommended quantization. Once it loads reliably, use the benchmark view and live memory readout to decide whether a larger model or context window fits your machine.