> ## Documentation Index
> Fetch the complete documentation index at: https://arka-agent-mintlify-f21cea86.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.

# Apple Intelligence (on-device)

> Use Apple's Foundation Models via Arka on supported Macs.

Arka can route LLM requests to Apple's on-device **Foundation Models** (Apple Intelligence) through the `apple-fm` provider. Inference stays on your Mac, and no API key is required.

## Requirements

* **macOS 26+** (Apple Intelligence enabled in System Settings)
* **Apple Silicon Mac** that supports Apple Intelligence
* Python optional extra: `pip install 'arka-agent[apple-fm]'` (installs `apple-fm-sdk`)

On Linux and Windows, the provider is registered but always skipped gracefully.

## Quick setup

```bash theme={null}
pip install 'arka-agent[apple-fm]'
arka model setup apple-fm
arka model status
arka llm apple-fm status
```

Set Apple Intelligence as your preferred provider:

```bash theme={null}
arka provider set apple-fm --model apple-fm-system
```

Or add to `.env`:

```bash theme={null}
APPLE_FM_ENABLED=1
AI_PREFERRED_PROVIDER=apple-fm
AI_PREFERRED_MODEL=apple-fm-system
```

## How it works

1. **Native SDK** — Arka calls `apple-fm-sdk` directly (`SystemLanguageModel` + `LanguageModelSession`).
2. **Fallback chain** — On supported Macs, `apple-fm` is added after Ollama in the auto-built fallback chain.
3. **Streaming** — Supported via the SDK's `stream_response()` when available.

Check availability anytime:

```bash theme={null}
arka llm apple-fm status
arka model status
```

## Optional: apple-fm-cli server

If the SDK is not installed but you run an OpenAI-compatible local server (for example [apple-fm-cli](https://pypi.org/project/apple-fm-cli/)), Arka auto-detects it:

```bash theme={null}
pip install apple-fm-cli
apple-fm-cli server --host 127.0.0.1 --port 8765
```

Configure Arka to use that endpoint (default port `8765` avoids conflict with vLLM on `8000`):

```bash theme={null}
export APPLE_FM_CLI_BASE_URL=http://127.0.0.1:8765/v1
export APPLE_FM_ENABLED=1
```

The server does not need a real API key; Arka sends a placeholder bearer token.

## Troubleshooting

| Symptom | Fix |
| - | - |
| `SDK not installed` | `pip install apple-fm-sdk` or `pip install 'arka-agent[apple-fm]'` |
| `Apple Intelligence not enabled` | System Settings → Apple Intelligence → turn on |
| `DEVICE_NOT_ELIGIBLE` | Mac or OS version does not support on-device models |
| `MODEL_NOT_READY` | Wait for Apple Intelligence model download to finish |

Disable the provider without uninstalling:

```bash theme={null}
export APPLE_FM_ENABLED=0
```

## Related

* [Model hosting setup](/guides/model-hosting) — Ollama, vLLM, LM Studio, Exo
* `arka llm providers --models` — list configured providers
* `arka hybrid status` — combine local and hosted routes


## Related topics

- [Stock intelligence](/guides/stocks.md)
- [AI agent guide — use Arka over MCP](/guides/ai-agents.md)
- [Skills catalog: 70+ built-in Arka commands](/guides/skills.md)
- [MCP integration for Cursor and Claude](/guides/mcp.md)
- [Hosted mode](/guides/hosted-mode.md)


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.