Complete Ollama keyboard shortcuts and commands reference — 29 shortcuts across 4 categories. Quick reference cheat sheet for Windows & Mac.
| Shortcut | Action |
|---|---|
| ollama pull <model> | Download model |
| ollama pull <model>:<tag> | Pull specific version |
| ollama list | List local models |
| ollama rm <model> | Remove model |
| ollama show <model> | Model info |
| ollama cp <src> <dst> | Copy model |
| ollama push <model> | Publish model |
| Shortcut | Action |
|---|---|
| ollama run <model> | Run interactive |
| ollama run <model> "prompt" | Single prompt |
| cat file.txt | ollama run <model> | Pipe input |
| /bye | Exit chat |
| /clear | Clear history |
| /set system <text> | Set system prompt |
| /show info | Model details |
| /show modelfile | Show Modelfile |
| Shortcut | Action |
|---|---|
| ollama serve | Start server |
| OLLAMA_HOST=0.0.0.0 ollama serve | Expose server |
| curl localhost:11434/api/generate | Generate API |
| curl localhost:11434/api/chat | Chat API |
| curl localhost:11434/api/tags | List models API |
| OLLAMA_MODELS=<path> | Custom model dir |
| OLLAMA_NUM_PARALLEL=4 | Parallel requests |
| OLLAMA_MAX_LOADED_MODELS=2 | Max loaded models |
| Shortcut | Action |
|---|---|
| FROM <model> | Base model |
| SYSTEM "<text>" | System prompt |
| PARAMETER temperature 0.7 | Temperature |
| PARAMETER num_ctx 4096 | Context length |
| TEMPLATE "{{ .Prompt }}" | Prompt template |
| ollama create <name> -f Modelfile | Build model |
Day-to-day Ollama is three commands: ollama pull <model> downloads, ollama run <model> opens an interactive chat, and ollama list shows what is on disk (ollama rm <model> reclaims the space — local models are gigabytes each). Tags select variants: ollama pull <model>:<tag> pins a size or quantisation instead of the default. For scripting, pass the prompt inline with ollama run <model> "prompt" or pipe a file straight in: cat file.txt | ollama run <model>.
Inside an interactive session, slash-commands control the chat: /clear wipes context, /set system <text> changes the system prompt on the fly, /show info and /show modelfile reveal what you are actually running, and /bye exits.
ollama serve runs the API server that everything talks to on port 11434 — curl localhost:11434/api/generate and curl localhost:11434/api/chat are the two endpoints most integrations use. Environment variables tune it: OLLAMA_MODELS=<path> relocates the model store to a bigger drive, OLLAMA_NUM_PARALLEL=4 allows concurrent requests, and OLLAMA_HOST=0.0.0.0 ollama serve exposes the server to your network. To make a customised assistant permanent, write a Modelfile — FROM <model>, a SYSTEM "<text>" prompt, PARAMETER temperature 0.7 and PARAMETER num_ctx 4096 — then build it once with ollama create <name> -f Modelfile and run it by name forever after.
Variants of the same model: parameter sizes and quantisation levels. Pulling without a tag gets the default; ollama pull <model>:<tag> pins an exact one. ollama show <model> reports what a downloaded copy actually is — worth checking before comparing quality or speed.
Treat it carefully. It binds the server to every network interface with no authentication of its own, so anyone who can reach the port can run your models. Fine on a trusted LAN; behind anything less, keep the default localhost binding or put a reverse proxy with auth in front.
In Ollama's model directory under your user profile by default. Set OLLAMA_MODELS=<path> (for the server process) to store them elsewhere — a common fix when the system drive is small, since a handful of models easily consumes tens of gigabytes.
Usually not on a desktop install, where the background service starts automatically and ollama run just works. Run ollama serve manually on headless machines, in containers, or whenever you want to control the environment variables the server starts with.
Write a Modelfile: FROM <model> for the base, SYSTEM "<text>" for standing instructions, PARAMETER lines for behaviour. Then ollama create <name> -f Modelfile builds it as a named local model — your settings persist without repeating /set system every session.
Open your assistant with this page preloaded as the source — great for follow-up questions like "which of these work in other apps?"