Skip to main content

Frequently Asked Questions

Everything you need to know about local AI on your Mac.

General
MacAI HK is a professional AI configuration service for Mac. We install and optimize open-source AI tools to run entirely on your local hardware, giving you complete privacy, zero subscriptions, and unlimited potential. Whether you bring your own Mac or purchase a pre-configured bundle, we handle the entire setup so you can focus on using AI, not configuring it.
BYOM stands for "Bring Your Own Mac." It means you provide your existing Apple Silicon Mac and we upgrade it with a full suite of AI capabilities. This is our most popular service because it lets you unlock the AI potential of hardware you already own, without purchasing a new machine.
No. Once setup is complete, all AI models run 100% locally on your Mac. You do not need an internet connection to use any of the AI tools we install. Internet is only required during the initial setup and for downloading additional models later if you choose to expand your collection.
No. MacAI HK is an independent AI configuration service and is not affiliated with, endorsed by, or authorized by Apple Inc. Mac, Mac mini, Mac Studio, and macOS are trademarks of Apple Inc. We are a service company that specializes in configuring open-source AI software on Apple hardware.
We provide services and documentation in English and Traditional Chinese. Our website is fully bilingual, and our team can assist you in either language. The AI models we install also support multiple languages depending on the model.
Hardware & Compatibility
Any Mac with Apple Silicon (M1 or later) is compatible. This includes MacBook Air, MacBook Pro, Mac mini, Mac Studio, Mac Pro, and iMac models from late 2020 onward. Minimum RAM is 8GB, but we recommend 16GB or more for smooth daily use, and 32GB or more for enterprise-grade AI workloads.
Yes. Any Apple Silicon Mac works with our service. MacBook Air, MacBook Pro, Mac mini, Mac Studio, Mac Pro, and iMac are all fully supported. The key requirement is Apple Silicon (M1 chip or later), not the form factor.
8GB of unified memory can run smaller AI models and is sufficient for basic tasks. 16GB is recommended for comfortable daily use with mid-size language models and image generation. 32GB or more is ideal for advanced workloads like running large language models, multiple models simultaneously, or enterprise-grade AI pipelines.
No. We only install software on your Mac. There are no hardware modifications of any kind. All tools we install are standard macOS-compatible applications and open-source software. Your Apple warranty and AppleCare coverage remain completely unaffected.
AI Models & Software
Depending on your service tier and hardware capabilities, we install models such as Llama 3.1, DeepSeek V3, Mistral, Qwen, Stable Diffusion for image generation, Whisper for speech-to-text, and more. The exact selection is tailored to your Mac's specifications and your intended use cases.
Yes. We install Ollama as your primary model management tool, which makes it simple to download, update, and manage additional AI models at any time. A single command is all it takes to add a new model. We also provide documentation on how to find and install models that suit your needs.
The AI suite we configure enables a wide range of capabilities: text generation and conversation, code assistance and generation, image creation with Stable Diffusion, speech-to-text transcription with Whisper, document question-and-answer with PrivateGPT, workflow automation with n8n, and much more. The specific capabilities depend on your service tier.
For many tasks, modern local models perform comparably to cloud-based AI services. They are especially strong for privacy-sensitive work, coding assistance, and document analysis. The key advantages of local AI are complete data privacy, no recurring subscription fees, no usage limits, and the ability to work offline. For specialized enterprise needs, local models can even be fine-tuned on your own data.
Service & Support
In-person BYOM setup typically takes 2 to 4 hours. Remote BYOM setup takes 2 to 6 hours depending on your internet speed and the number of models being downloaded. Pre-configured Mac bundles require 3 to 7 business days for preparation and delivery.
Yes. We offer remote setup via secure remote access tools for both international customers and Hong Kong customers who prefer not to visit in person. The remote session is conducted in real time, and we guide you through every step of granting temporary access so that your data remains protected throughout the process.
Every service tier includes complimentary configuration support ranging from 14 to 90 days after setup, depending on your tier. During this period we help with any questions, troubleshooting, or adjustments. After the included period, optional monthly support plans are available starting from $49/month for ongoing assistance, model updates, and priority support.
We offer a full refund if you cancel before the service begins. Once the AI configuration has been completed on your Mac, refunds are not available as the service has been fully delivered. Please see our Terms of Service for complete details on our refund policy.
Simply email us at [email protected] or click the "Book Free Assessment" button on any page of our website. We will assess your Mac's hardware, discuss your AI use cases, and recommend the best service tier for your needs — all at no cost and no obligation.

Still have questions?

Get a free, no-obligation assessment of your Mac's AI potential.

Book Free Assessment