gollama

gollama is a terminal app for people managing a local LLM collection in Ollama. Its keyboard-driven interface helps you compare downloaded models and clear out ones you don't need.

• Model inspection and cleanup: Search, filter and sort your collection by size, quantisation, model family or modification date, inspect individual models and select models for deletion.
• Model configuration: Copy or rename models and edit their Ollama Modelfiles to change configuration.
• Running models and updates: See which models are running, launch or unload them, pull updates while preserving your configuration and push models to a registry.
• Memory estimates: Estimate RAM or VRAM needs for downloaded Ollama models or Hugging Face model IDs, compare quantisation and context settings, and find settings that fit a memory budget.

The app runs on macOS and Linux and is open source under the MIT license. It manages Ollama models; LM Studio linking isn't available.

Memory estimates are approximate and include different key/value cache settings. The estimator needs an internet connection to fetch model configuration from Hugging Face when that information isn't cached locally.