Player not loading? Watch on YouTube
This tutorial explains how to connect Hermes Agent to LM Studio as a local model provider. The speaker presents the setup as a way to use Hermes offline and walks through model selection, starting the server, and choosing the provider in Hermes.
The setup begins in LM Studio's model browser. The example uses a model the speaker calls Gemma 4, but the speaker advises against using that particular version at the end. LM Studio displays warnings when models are likely too large for the machine. The tutorial also discusses quantized versions as lighter options, so the model choice depends on the user's hardware rather than the example alone.
With the example model still downloading earlier in the walkthrough, the speaker opens the local server panel, toggles the server on, and explains loading a model. In the terminal, the speaker opens Hermes setup and selects LM Studio from the provider options. Once Hermes chat starts, typing "model" opens the model selection flow, where the user can switch to LM Studio and choose the loaded model. The interface reports the context window, provider, and selected model.
Ollama appears as another local provider, but the supplied transcript ends before a separate Ollama walkthrough. The tutorial covers the connection steps for a local AI agent; it does not demonstrate a completed agent task or test the setup with the network disconnected.
