Download a local model

Download a local model

Use Model Manager’s built-in Hugging Face browser when you want LM Nexus to download a GGUF and create the initial profile for it.

Before you start

You need:

• a running LM Nexus private-alpha/source environment;
• access to Model Manager;
• enough disk space for the GGUF you choose;
• network access to Hugging Face.

1. Open Model Manager

Open Model Manager.

The left side panel contains:

• Profiles
• Library
• Groups
• Downloads
• API Serving

2. Open the Hugging Face browser

In the Model Manager toolbar, choose Browse HuggingFace.

The browser lets you search/filter repositories, open repository details, and select downloadable GGUF files.

3. Choose the GGUF

Open the model repository you want and select the GGUF file you intend to use.

Quantization and model size affect RAM/VRAM requirements. If you are unsure, start with a model that comfortably fits your available hardware rather than the largest option.

4. Start the download

Choose Download selected.

LM Nexus starts a download job for the selected GGUF.

5. Watch progress

Open the Downloads section to follow active jobs and download history.

When the download completes, the GGUF becomes part of the local Library.

Expected result

When the download finishes, the GGUF appears in Model Manager’s local library.

LM Nexus also creates a default profile for the downloaded GGUF, so it should appear under Profiles without extra setup.

If it does not appear

Try:

1. open Downloads and confirm the job actually completed;
2. choose Refresh in Model Manager;
3. choose Rescan if the file exists locally but is not discovered;
4. check the Model Manager logs for the download or discovery error.

See Troubleshooting if the problem is broader than one download.

Next

• Manage a model profile
• Start or stop a model