Your Mac might already have a free AI chatbot built right into it, and you probably had no idea it was there. No ChatGPT subscription, no downloading a separate AI app and no internet connection required once the model is ready to go. Just local AI when you need it.
That’s what X user Alex Choi discovered this week when he posted about finding a local large language model (LLM) hiding inside his MacBook’s Terminal app.
“wtf i just discovered that a macbook comes with a free local LLM in your terminal,” he wrote in a post on October 21 that attracted more than 1 million views in the accompanying screenshot.
The discovery caught people’s attention for a good reason. We’re used to opening ChatGPT, Claude or Gemini in a browser whenever we need AI assistance. But if you’re running macOS 27 on a compatible Mac, Apple has already given you another option. And accessing it is surprisingly easy.
Your Mac has its own AI chatbot
The feature is called Apple Foundation Models CLI, or fm for short. Apple introduced it with macOS 27 Golden Gate, which arrived in September.
Unlike ChatGPT or Claude, which typically send requests to cloud servers, Apple’s built-in tool can run an AI model directly on your Mac. This means a you can ask questions, brainstorm ideas, rewrite text and get help with everyday tasks without sending your prompts to an external AI service.
Apple demonstrated the feature during WWDC 2026, explaining that the command-line tool comes preinstalled with macOS 27.
It’s primarily designed for developers, but there’s no reason everyday Mac users can’t experiment with it, too. In fact, because it’s local AI, it’s a great way to learn more about what AI can do while keeping your information out of the cloud.
The catch? You’ll need to open Terminal, which might look intimidating if you’ve never used it. Fortunately, getting started only requires a couple of commands.
How to access your Mac’s free AI chatbot
First, make sure you’re running macOS 27 and using a Mac with Apple silicon. You’ll also need Apple Intelligence enabled and its on-device models downloaded.
Here’s how to get started.
1. Open Terminal. Press Command + Space, type “Terminal” and press Return.
2. Check that Apple’s AI tool is available. Type fm and press Return. If prompted to accept Apple’s Foundation Models CLI terms, you’ll need to review and accept them before continuing. You can do this by entering sudo fm license, providing your Mac administrator password and following the prompts.
3. Start chatting. Once the tool is ready, type fm chat and press Return.
4. Ask a question. You should now see an interactive interface where you can enter a prompt and receive a response from Apple’s AI model.
That’s it. From there, you’re chatting with an AI model running locally on your Mac. You can start with something simple, such as “Give me five easy dinner ideas using chicken and rice” or “Explain compound interest like I’m 12.”
You can also ask follow-up questions, just as you would with other AI chatbots.
If you don’t want an ongoing conversation, there’s another useful command. Type fm respond followed by your question in quotation marks to get a single response without opening the chat interface.
What can you actually do with it?
Apple’s local AI isn’t designed to replace the most advanced versions of ChatGPT, Claude or Gemini. It’s a smaller model intended for tasks that don’t necessarily need a massive cloud-based AI system.
You could ask it to rewrite a long email so it sounds more professional, brainstorm birthday party themes, explain an unfamiliar term or help outline a project.
One of the biggest advantages is privacy. Because the on-device model processes requests locally, your prompts don’t need to be sent to an outside server to generate a response. That also makes it useful when you’re traveling or working somewhere with unreliable Wi-Fi.
There’s even another benefit: cost. You don’t need an additional AI subscription, and there are no per-message charges for using the local model through the built-in tool.
Of course, there are limitations. Smaller models can struggle with complicated reasoning, lengthy conversations and questions that require current information. And because the model runs locally, it doesn’t automatically have access to the latest news or information from the internet.
There’s one catch: Not every Mac supports it. So, before you rush to open Terminal, there are a few requirements worth knowing. Apple’s fm command-line tool comes with macOS 27, so it won’t be available as a built-in feature on Macs running older versions of macOS.
You’ll also need compatible hardware. Apple Intelligence supports Macs with M1 chips or later, along with the Apple silicon-equipped MacBook Neo. Older Intel-based Macs aren’t supported.
Apple Intelligencemust also be enabled, and the necessary on-device models must finish downloading. You can check your settings under System Settings and Apple Intelligence & Siri.
If you’re unsure whether your Mac qualifies, Apple has a full list of Apple Intelligence requirements.
The takeaway
There’s something refreshing about discovering that a computer you already own can do something you previously relied on a separate AI service for.
Apple has been building AI into macOS for years, but this little Terminal trick makes that technology feel more accessible, and the best part is that you don’t have to pay anything extra to try it.
- OpenAI unleashed thousands of AI agents to study a Parkinson-like disease — here’s what they found in 32 minutes
- 7 ChatGPT plugins that are a must-have for anyone with a side hustle
- Google just cut down 250,000 trees for AI data centers — and now it’s facing a major backlash
