← The Nexus
01 Current

Models & Runtime

Run local models without giving up control.

This is where you download models, choose how they run, and connect local, self-hosted, or external providers.

Models & Runtime Agents & Tools Workspace Physical Systems Living Worlds
LM Nexus Models & Runtime models · providers · runtime

What's here

What this area includes

  • local GGUF library and Hugging Face downloads
  • profiles, variants, groups and runtime settings
  • llama.cpp start/stop control, logs, and runtime metrics
  • OpenAI-compatible local serving plus self-hosted and connected providers

How it connects

How it fits with the rest

Agents and other parts of Nexus use the same model profiles and providers, while you can still see the runtime underneath.

Models & Runtime tool / context / capability other Nexus realms
In the app 01

What this looks like in LM Nexus

These screenshots come from the current private alpha. The interface is still evolving, but the workflows shown here are real.

LM Nexus Model Manager interface.
Current view Model Manager

Download models, save profiles, and change runtime settings from the same workspace.

LM Nexus runtime panel.
Current view Runtime panel

See whether the runtime is running, what it is doing, and how it was started.

Status

What's available today

LM Nexus can manage the runtime for you, or you can connect directly to the server when the runtime supports it.

Go deeper

Keep exploring

The Product Tour shows how the pieces connect. The docs go deeper when you want setup steps, controls, or technical detail.

Models →Runtimes →Download a model →Use the local API → Back to Product Tour →
Connected realms 01—05

The parts stay separate, but they can work together through tools and shared context.