warning

Legacy managed-runtime reference. Command examples below describe the older desktop/managed llama.cpp preview and are legacy, not the current BCCLI v0.4 command surface. Use BCCLI and Local AI for current instructions.

Installed models

List installed models

powershell

botconnector ls

Shows each downloaded model: repository id, quantization, file size, disk path and whether it is currently loaded.

In the desktop app

Installed lists the same store with one-click Run, Unload and Delete. The CLI and GUI share the store, so a model loaded in the GUI appears in botconnector ps.

Inspecting files

Each installed model keeps a metadata snapshot (Hub metadata at download time) plus its GGUF file(s). Multi-part shards are reassembled into one logical model entry.

powershell

botconnector ls --json

returns structured output for scripts — see botconnector models.

Loading and unloading

powershell

botconnector run <publisher>/<model>
botconnector unload

Details: Run a model. One model occupies memory at a time in the preview; switching models unloads the previous one.

Removing a model

Desktop: Installed → Delete. CLI:

powershell

botconnector models remove <publisher>/<model>

Removal deletes the local files only — the model remains on Hugging Face and can be re-downloaded.

Re-download / repair

If a file is corrupted (checksum mismatch at load), delete and download again:

powershell

botconnector get <publisher>/<model>@Q4_K_M

Troubleshooting

Missing files, wrong sizes, corrupted downloads: Download problems.