Local models pile up fast and each one is gigabytes. Here's the handful of Ollama commands to list, update, remove, and inspect your models — and keep your disk under control.
With a handful of simple [Ollama](/ollama-vs-lm-studio-which-should-you-use-2026) commands. Local models are convenient but they pile up fast, and each one is several gigabytes, so a little housekeeping keeps your disk under control and your setup tidy. The essentials: `ollama list` shows what you have, `ollama pull <model>` downloads or updates one, `ollama rm <model>` removes one you don't use, and `ollama ps` shows what's currently loaded in memory. That's most of what you'll ever need. Here's the full toolkit for keeping your local AI organized.
The commands you'll actually use
Ollama's model management is refreshingly simple — a few commands cover everything. Here's the practical set:
ollama list # see all your models and their sizes
ollama pull qwen3:8b # download OR update a model (pull again = latest)
ollama rm llama3.1:8b # delete a model you don't use (frees GB)
ollama ps # see what's loaded in memory right now
ollama show qwen3:8b # inspect a model's details
ollama cp qwen3:8b my-qwen # copy a model (e.g. before customizing it)
The two you'll run most are `ollama list` (to see what's taking up space) and `ollama rm` (to clear out models you tried once and don't use). Since models are gigabytes each, a quick ollama list every so often, followed by removing the ones you don't need, keeps your disk healthy.
A handful of Ollama commands — list, pull, rm, ps — is all you need to keep your local models tidy. · Unsplash
Updating models and reclaiming disk
Two bits of housekeeping matter most. Updating a model: Ollama models get improved versions over time, and to update, you simply `ollama pull <model>` again — it fetches the latest version of that tag. There's no separate 'update' command; a re-pull is the update. It's worth doing occasionally for models you rely on, since newer versions often fix bugs or improve quality. Reclaiming disk space: this is the one that bites people. Every model you pull lives in the `~/.ollama` folder (or the equivalent on Windows), and at several gigabytes each, a collection of models you were 'just trying out' can quietly consume tens of gigabytes. Run `ollama list` to see the damage, then `ollama rm <model>` on the ones you don't use — that immediately frees the space. If you're tight on disk, keep just your two or three go-to models installed and remove the rest; you can always re-pull one in a couple of minutes when you need it. That simple discipline — list, remove the unused, keep the essentials — is all it takes to run a tidy local-AI setup indefinitely. Pair it with sizing models to your VRAM and you'll never be surprised by a full disk or a model that won't fit.
Quick answers
How do I update a model in Ollama?
Just pull it again: run 'ollama pull <model>' (for example, 'ollama pull qwen3:8b'), and Ollama fetches the latest version of that model tag. There's no separate update command — re-pulling is the update. It's worth doing occasionally for models you use regularly, since newer versions often fix bugs or improve quality. To check what you have first, run 'ollama list', which shows all your downloaded models. Updating won't create a duplicate; it replaces the model with the current version under the same tag.
How do I free up disk space from Ollama models?
Run 'ollama list' to see all your downloaded models and their sizes, then 'ollama rm <model>' to delete the ones you don't use — this immediately frees the space. Ollama models are several gigabytes each and live in the ~/.ollama folder (or the equivalent on Windows), so a collection of models you tried once can quietly consume tens of gigabytes. Keep just your two or three go-to models installed and remove the rest; you can always re-pull a model in a couple of minutes when you need it again.
What are the essential Ollama commands?
The core commands are: 'ollama list' (see your downloaded models and sizes), 'ollama pull <model>' (download or update a model), 'ollama run <model>' (chat with a model), 'ollama rm <model>' (delete a model to free disk space), 'ollama ps' (see which models are loaded in memory), and 'ollama show <model>' (inspect a model's details). These cover essentially all day-to-day model management. The two you'll use most for housekeeping are 'ollama list' and 'ollama rm' to keep your disk usage under control.
