Repolode
Submit
sammcj

sammcj/gollama

A macOS/Linux terminal UI for listing, inspecting, copying, pushing, and deleting locally installed Ollama models.

View on GitHubWebsite
Stars
1,843
Forks
112
Open issues
0
Last commit
17 days ago

What is it?

Gollama is a text user interface (TUI) for managing models in Ollama. It lists installed models with metadata such as size, quantisation level, family and modified date, and lets you sort, filter, select and act on them with hotkeys. Actions include running and unloading models, inspecting details, editing a model's Modelfile, copying or renaming models, pushing them to a registry, and viewing currently running models.

Beyond the interactive interface, it offers command-line flags for listing, searching with OR and AND patterns, editing Modelfiles, and unloading all running models. A vRAM estimator calculates memory needs for an Ollama model or a Hugging Face model across quantisation levels and context lengths, and can suggest the best quantisation for a given memory budget.

Configuration lives in a JSON file under ~/.config/gollama, covering default sort order, visible columns, API host, log level, an editor for Modelfiles, and custom themes. The project is written in Go and installs with go install.

Key features

  • 01Interactive TUI with hotkeys for running, deleting, copying, and pushing models
  • 02Sort and filter models by name, size, date, quantisation, family, or parameter size
  • 03Inspect model details and edit Modelfiles in your preferred editor
  • 04Estimate vRAM usage for Ollama or Hugging Face models across quantisation and context settings
  • 05Command-line mode for listing, searching, and unloading models

Use cases

  • Cleaning up old or unused Ollama models that take up disk space
  • Checking which models are loaded and unloading them to free memory
  • Estimating whether a model will fit in available GPU or system memory before pulling it

Quick start

  1. 1

    Install with Go (recommended)

    go install github.com/sammcj/gollama/v2@latest
  2. 2

    Run gollama

    gollama

Things to know

  • Supports only macOS and Linux, not Windows.
  • The README says development has slowed and the author rarely uses Ollama anymore.
  • LM Studio linking was removed in v2.0.1.

In collections

Terminal UI appsRun LLMs locally

Topics

#ollama#llm#terminal-ui#model-management#go#vram#ai#ggml#gguf#linux

FAQ

What is gollama?
A macOS/Linux terminal UI for listing, inspecting, copying, pushing, and deleting locally installed Ollama models. Gollama is a text user interface (TUI) for managing models in Ollama. It lists installed models with metadata such as size, quantisation level, family and modified date, and lets you sort, filter, select and act on them with hotkeys. Actions include running and unloading models, inspecting details, editing a model's Modelfile, copying or renaming models, pushing them to a registry, and viewing currently running models.
Is gollama free to use? What license does it have?
gollama is open source under the MIT license. Check the license terms before using it in commercial products.
Is gollama still maintained?
Actively maintained. The last commit to gollama was on Oct 21, 2026.
What can I use gollama for?
Cleaning up old or unused Ollama models that take up disk space Checking which models are loaded and unloading them to free memory Estimating whether a model will fit in available GPU or system memory before pulling it
How do I install gollama?
Install with Go (recommended): go install github.com/sammcj/gollama/v2@latest Run gollama: gollama
Where can I get gollama?
The source code and installation instructions are on GitHub: https://github.com/sammcj/gollama

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects