GGUF Import Guide#

Use this flow when you already have local model files and want to import them into the app.

Supported inputs#

The Import Model flow accepts:

  • .gguf file only

  • Modelfile only

  • Modelfile + .gguf together

Important rule:

  • If your Modelfile references a GGUF in its FROM line, import both files together in one action.

Open the importer#

  1. Open Marketplace.

  2. Select Import Model.

  3. Use drag-and-drop or click upload in the modal.

Choose import target host#

In the importer modal, set Host to:

  • Local for your local runtime

  • A configured remote host if you want the import on that remote runtime

File handling behavior#

  • The app validates file types and shows errors for unsupported files.

  • If a Modelfile is present, the app shows a parsed preview.

  • If a referenced GGUF is missing, import is blocked with a clear status message.

Optional alias#

You can set an alias (up to 20 characters) before import. This is used as the model name for the imported artifact.

Import scenarios#

GGUF only:

  1. Select one .gguf file.

  2. Set optional alias and host.

  3. Click Import.

Modelfile only:

  1. Select one Modelfile.

  2. Set optional alias and host.

  3. Click Import.

Modelfile + GGUF:

  1. Select both Modelfile and its referenced .gguf.

  2. Confirm the status shows the reference is resolved.

  3. Set optional alias and host.

  4. Click Import.

After import#

  1. Open Chat.

  2. Select the same host you imported to.

  3. Pick the imported model and run a test prompt.

Common failures#

  • Only .gguf or Modelfile allowed: remove unsupported file types.

  • Modelfile references .gguf: add the referenced GGUF file in the same import action.

  • Import fails on remote host: verify remote host connectivity and permissions.