Skip to content
Stork.AIstork.ai
AI Tool

apfel Review

Apfel unlocks your Mac's built-in AI via CLI and an OpenAI-compatible server, offering free, private, on-device inference.

shipped Jul 7, 2026aifreemium
ai
apfel — product screenshot

Why it matters

1Apfel is a free, open-source command-line tool for macOS Tahoe (version 26) and later.
2It provides access to Apple's on-device 3-billion-parameter LLM without API keys or subscriptions.
3The tool runs 100% on-device, ensuring data privacy and offline functionality.
4Installation is streamlined via a single `brew install` command (v1.0.0 as of April 13, 2026).

About apfel

Business Model
Open Source
Headquarters
N/A
Founded
2026
Team Size
N/A
Funding
Bootstrapped
Platforms
macOS
Target Audience
Apple Silicon users

Pricing Plans

Free
$0 / lifetime
  • • No token costs
  • • No API keys
  • • No subscriptions
  • • On-device AI

Leadership

Franz EnzenhoferFounderLinkedIn
Arthur FicialCo-maintainerLinkedIn
GitHubOpen Source

Specs

API Docs

View Documentation →

GitHub

View Repository →

API Available

Yes, public API

overview

What is apfel?

apfel is an on-device AI access tool developed by Arthur-Ficial that enables Apple Silicon Mac users to leverage Apple's built-in Foundation Models. It offers a command-line interface (CLI) and an OpenAI-compatible HTTP server, enabling users to interact with the approximately 3-billion-parameter Large Language Model (LLM) integrated into macOS Tahoe (version 26) without requiring API keys, subscriptions, or cloud processing. Launched around April 2026, Apfel provides direct, local access to Apple Intelligence capabilities for various tasks, including text summarization, code explanation, and document Q&A, all processed entirely on the user's Mac.

features

Key Features of apfel

Apfel provides a suite of features designed to maximize the utility of Apple's on-device AI models, focusing on privacy, ease of use, and developer compatibility.

  • Zero configuration required for immediate use after installation.
  • Runs entirely on-device, ensuring no data leaves the user's Mac.
  • Offers an OpenAI-compatible local server for seamless integration with existing tools and development workflows.
  • Includes a multi-turn conversational chat interface for interactive AI interactions.
  • Provides direct command-line interface (CLI) access for scripting and automation.
  • Supports natural language processing tasks such as summarization and rewriting.
  • Enables code explanation and short coding tasks directly from the terminal.

use cases

Who Should Use apfel?

Apfel is designed for a diverse range of users who prioritize privacy, local processing, and direct access to on-device AI capabilities on Apple Silicon Macs.

  • Privacy-conscious professionals: Lawyers, healthcare workers, and journalists can handle sensitive topics offline without data leaving their machine.
  • Writers and content creators: Utilize on-device AI for drafting, editing, rewriting text, and generating summaries.
  • Developers: Explore Apple's ML frameworks, build local tools using the OpenAI-compatible server, and perform code explanation or short coding tasks.
  • Students: Access AI assistance for quick questions, explanations, and lookups without requiring subscriptions.
  • Mac power users: Integrate AI into their daily terminal automation and customize workflows with CLI access.

how to use

How to Use apfel

Getting started with Apfel is designed to be straightforward, leveraging macOS's package manager for quick installation and immediate access to the on-device AI.

  • 1Ensure your Mac is running macOS 26 (Tahoe) or later and has an Apple Silicon chip (M1 or newer).
  • 2Open your terminal application.
  • 3Install Apfel using Homebrew: brew install apfel.
  • 4Start an interactive chat session: apfel chat.
  • 5Run a single prompt: apfel 'Explain the concept of quantum entanglement.'.
  • 6Start the OpenAI-compatible local server: apfel server.

pricing

apfel Pricing & Plans

Apfel operates on a freemium model, with its core functionality being entirely free and open-source. There are no subscription fees, API keys, or per-token costs associated with using the on-device AI model. All processing occurs locally on the user's Apple Silicon Mac, eliminating traditional cloud-based pricing structures. The project is maintained by Arthur-Ficial and is available for free on GitHub.

  • Free: $0 (lifetime access to all features, 100% on-device processing)

Pros

  • +Completely free and open-source, with no hidden costs or subscriptions.
  • +Ensures 100% data privacy as all processing occurs on-device, with no data leaving the Mac.
  • +Zero configuration and easy installation via brew install on macOS Tahoe (version 26) and later.
  • +Provides an OpenAI-compatible local server, facilitating local development and tool integration.
  • +Offers fast response times for quick queries on Apple Silicon Macs due to on-device inference.
  • +Integrates well into terminal workflows with CLI access, JSON output, and pipe support.

Cons

  • −Limited context window of 4096 tokens (input + output combined), restricting utility for long conversations or complex documents.
  • −The underlying Apple Foundation Model (3B parameters) is less powerful than frontier models like GPT-4 or Claude for complex reasoning tasks.
  • −Requires macOS 26 (Tahoe) or later and an Apple Silicon Mac (M1 or newer), limiting compatibility.
  • −While the underlying model supports multiple languages, optimal performance is primarily observed in English.
  • −Lacks the polished graphical user interface and extensive feature breadth of some commercial Mac AI utilities.

Policies

Free Tier

Vendor website advertises a free tier.

Similar Tools

apfel vs Competitors

Apfel occupies a unique niche by leveraging Apple's pre-installed on-device AI, differentiating it from other local LLM runners and cloud-based services.

1
Ollama↗

Ollama simplifies running a wide range of open-source large language models locally with a command-line interface and an OpenAI-compatible API.

Unlike Apfel, which leverages Apple's built-in on-device model, Ollama requires users to download models (often several gigabytes) from its extensive library. Both are open-source, CLI-focused, and provide an OpenAI-compatible local server for running AI models.

2
LM StudioOn StorkCompare

LM Studio provides a user-friendly desktop application with a graphical interface for browsing, downloading, and interacting with local LLMs, including an OpenAI-compatible local server.

LM Studio offers a GUI-first experience for local AI, contrasting with Apfel's CLI-centric approach. Both provide an OpenAI-compatible local server, but LM Studio focuses on a wider range of downloadable models rather than exclusively Apple's on-device model.

3
LocalAIOn StorkCompare

LocalAI is an open-source, OpenAI-compatible AI engine that allows running various models (LLMs, vision, voice, image) locally on diverse hardware.

LocalAI is a more comprehensive open-source AI engine, acting as a drop-in replacement for the OpenAI API for various model types, whereas Apfel specifically targets Apple's on-device LLM. Both are open-source and offer OpenAI API compatibility for local inference.

4
JanOn StorkCompare

Jan is an open-source desktop application that provides a ChatGPT-like chat interface for running local large language models offline.

Jan offers a polished, familiar chat interface for local LLMs, similar to Apfel's chat mode, but it focuses on downloading and managing a variety of open-source models rather than exclusively leveraging Apple's built-in model. Both are open-source and free.

Visit apfel↗
Connect
⌘
GitHubgithub.com/Arthur-Ficial/apfel

AI Reputation Report

Is apfel yours?

ChatGPT, Perplexity, Gemini, Claude & Grok answer buyer questions about apfel every day. See whether they name apfel — or send buyers to a rival.

See what AI saysfree preview

Featured in articles