Free & Open Source

Apple Intelligence,
called by your LLM.

An MCP server that connects Claude and local LLMs to Apple's native on-device frameworks. OCR, PDF reading, entity detection, summarization, classification — all on your machine. No cloud. No API keys.

Zero tokens consumed Zero data leaves your Mac Reads PDFs & images Apple Silicon Claude Desktop + local LLMs macOS 26+

Eight tools. All on-device.

When your agent calls any of these, the processing runs through Apple's native frameworks — not through tokens, not through a cloud API.

`ocr`
No Apple Intelligence

Apple Vision OCR on any image or PDF. Hand it a PDF and it rasterizes and reads every page (or one via `page`). PNG, JPG, HEIC, TIFF, PDF. ~0.2s/page.

`pdf_text`
No Apple Intelligence

Pull a PDF's embedded text layer instantly via PDFKit — no OCR, exact, sub-second. The fast path for digital PDFs; falls back to `ocr` for scans.

`detect`
No Apple Intelligence

NSDataDetector extracts phone numbers, URLs, dates, and addresses from freeform text. Deterministic — no model involved.

`extract`
Apple Intelligence

Structured entity extraction from any text. Returns a JSON object. Optionally specify which fields you want.

`classify`
Apple Intelligence

Text classification against a label set you define. Returns a single label from your list.

`summarize`
Apple Intelligence

On-device summarization via Apple's FoundationModels framework. Fast, private, no tokens consumed.

`generate`
Apple Intelligence

On-device drafting, rewriting, and short-form answers. Accepts an optional system instruction.

`recognize_document`
No Apple Intelligence

Structured OCR via Apple Vision. Returns full transcript plus tables as rows and cells. Unlike `ocr`, layout survives — use this for charts, forms, invoices, and multi-column docs. Takes images or PDFs. (Requires macOS 26.)

Who it's for.

Built for anyone running agent workflows on a Mac who'd rather not route every document task through a cloud API.

Local LLM Users

Close the Apple Silicon gap

You're running Ollama, LM Studio, or a local Claude setup. Your Mac has serious on-device intelligence built in. This server connects the two.

Claude Desktop Users

Offload OCR and extraction without burning tokens

Route document parsing and entity detection to Apple Intelligence. The result comes back as tool output — no tokens consumed on those operations.

Privacy-First Workflows

Nothing leaves the machine

Every call — OCR, extraction, summarization — runs through Apple's on-device frameworks. Zero data leaves your Mac. Works fully air-gapped.

Developers

Add Apple Intelligence to any MCP-compatible client

Standard MCP server. Drop it into Claude Desktop, Voical, or any MCP-compatible local setup. Open source — fork it, extend it, or open a PR.

What you need.

Honest about limitations. Not every tool needs Apple Intelligence — but some do.

Apple SiliconM1 or later required for all tools
macOS 26+Required for `recognize_document` and the four Apple Intelligence tools (`extract`, `classify`, `summarize`, `generate`)
Any Apple Silicon Mac`ocr`, `pdf_text`, and `detect` run on any macOS — no Apple Intelligence, no macOS 26
MCP-compatible clientClaude Desktop, Voical, or any local MCP setup
No API keysNo cloud account, no subscription, no tokens

One honest note: if you're driving this from a cloud assistant (e.g. Claude Desktop), tool results return to that cloud conversation. For a fully air-gapped pipeline, drive it from a local client — Voical does this natively.

Two modes. One story.

Depending on your setup, there are two ways to run this.

Claude Desktop

Frontier reasoning + local processing

Claude handles the reasoning. Apple Intelligence handles the heavy lifting — OCR, extraction, classification — without consuming tokens. Tool results return to the cloud conversation.

Voical — End-to-end local

Nothing leaves your machine

Voical drives this MCP from a local model with persistent memory. No cloud conversation. No data anywhere but your Mac. The fully offline path. Learn more →

Both paths are valid. Pick based on whether frontier reasoning or complete data sovereignty matters more to you.

Use cases, combos & quick start.

Detailed use-case cards, tool chain combos, token savings, and setup Q&A.

Free. Open source. Install in 5 minutes.

If it's useful, a star helps. If it's broken for your setup, open an issue. That's how this gets better.

★ Star on GitHub →