> ## Documentation Index
> Fetch the complete documentation index at: https://docs.runanywhere.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Run a model locally

> Smaller models on your own machine, in three commands.

Most of what Wally runs is hosted — see the [Cloud quickstart](/wally-cloud/quickstart). When a
small model on your own machine is enough, the same CLI runs it, offline and without an
account.

## 1. Install

```bash theme={null}
curl -fsSL https://raw.githubusercontent.com/RunanywhereAI/wally/main/install.sh | sh
```

Check it landed:

```bash theme={null}
wally --version
```

## 2. Pull a model

```bash theme={null}
wally pull qwen3-0.6b
```

About 640 MB. `wally list --all` shows everything available, and `wally list` shows what you
already have.

## 3. Run it

```bash theme={null}
wally run qwen3-0.6b
```

That opens an interactive session. To ask one question and exit:

```bash theme={null}
wally run qwen3-0.6b "why is the sky blue"
```

## Where to go next

<CardGroup cols={2}>
  <Card title="Integrations" icon="plug" href="/integrations">
    Point opencode at the model you just pulled.
  </Card>

  <Card title="Wally Cloud" icon="cloud" href="/wally-cloud/index">
    Frontier-scale models we host, through the same commands.
  </Card>

  <Card title="On-device SDK" icon="mobile-screen" href="/sdks">
    Ship this inside your own app, on six platforms.
  </Card>

  <Card title="Serve it over HTTP" icon="server" href="/integrations#anything-else">
    `wally opencode --serve` prints a local OpenAI-compatible base URL for the model you pulled.
  </Card>
</CardGroup>