gollama

gollama is a terminal app for people managing a local LLM collection in Ollama on macOS or Linux. It puts model details and management actions in one keyboard-driven view, useful when you need to compare downloaded models or clear out ones you no longer use. It's open source under the MIT license.

The model list shows file size, parameter size, quantisation, family and modification date. You can search and filter the collection, sort it by those details, inspect individual models and select models for deletion. It also supports copying models and editing their Ollama Modelfiles.

Running models have their own view. You can launch a model or unload it from memory, pull new models, update existing ones while preserving user configuration, and push models to a registry. The app connects to an Ollama API on your machine or another host, so the models it manages live wherever that Ollama instance runs.

Its memory estimator helps you judge which model settings fit your hardware before committing to them. It accepts downloaded Ollama models and Hugging Face model IDs, estimates memory needs across quantisation levels and context lengths, and compares KV cache settings. It can also estimate the maximum context length for a memory budget or find a quantisation setting that fits a chosen context length. For Hugging Face estimates, it fetches model configuration over the internet when that information isn't cached locally.

The maintainer reports slower development. LM Studio linking was removed, so gollama should be evaluated as an Ollama manager.