What this does

Downloads a compact open-source language model from the Ollama library to the local machine and starts an interactive session. The model runs entirely offline after download and responds to text prompts without data leaving the system.

Steps

Pull the Llama 3.2 model. Downloads the 1-billion-parameter variant from the default library. Requires an active internet connection.
```
ollama pull llama3.2
```
Expected output: Progress bar followed by success.
Start an interactive inference session. Launches the model and exposes a command-line prompt.
```
ollama run llama3.2
```
Expected output: >>> Send a message (type /bye to exit).
Ask a simple question. Tests the model's ability to generate a coherent response.
```
>>> What is 2 + 2?
```
Expected output: A numbered greeting and a text response containing the answer "4".
Exit the session. Returns control to the shell.
```
>>> /bye
```
Expected output: Command prompt returns.

Verification

ollama list
# Expected: tabular list showing llama3.2 with a size in bytes and a MODIFIED timestamp

Common failures

no connection - Network is unavailable; check Wi-Fi or run pull again later.
model not found - Typo in model name; run ollama pull llama3.2 exactly as shown.
insufficient memory - Host has less RAM than the model requires; try a 1B-parameter variant or a smaller model.
permission denied - Sandbox forbids model compilation; launch Ollama directly from the host shell.
storage full - Disk space is exhausted; free at least 2 GB on the Ollama model directory volume.

Operator checkpoint

Before treating this as solved, write down the local runtime, model or package version, hardware/backend if relevant, and the verification output. This keeps the guide useful as a Will-It-Run style decision instead of a one-off command transcript.

How to pull and run your first open-source LLM using Ollama

What this does

Steps

Verification

Common failures

Operator checkpoint

Related guides