Skip to content

Choosing and switching AI models

Typeahead recommends a built-in model for your Mac. Here's how to switch to another one, or bring your own GGUF model file.

Last updated October 12, 2026

Typeahead's suggestions come from an AI model that runs on your Mac. Gemma and Qwen are built in, or you can bring your own. Every model is included on every plan.

The built-in models

During setup, Typeahead recommends the one that suits your Mac:

  • A Gemma model, the default. A strong all-rounder, and the best pick if you write in languages other than English.
  • A smaller Qwen model, for lighter Macs. It's English-first.
  • A larger Gemma model, for Apple silicon Macs with 24 GB of memory or more. It gives higher-quality suggestions.

Start with the one Typeahead recommends. Updates never change the model you picked.

Switch models

  1. Open Settings › Model.
  2. Pick a model from the list. A star marks the one recommended for your Mac.
  3. Click Download if you haven't downloaded it yet. Typeahead switches to it when the download finishes. For a model you already have, click Use This Model.

Each built-in model is a one-time download, so switch while you're online. The recommended model is about a 4 GB download, so check you have the space. An interrupted download picks up where it left off, and each file is checked before Typeahead uses it.

Bring your own model

Under Settings › Model › Custom Model, click Browse… and choose a GGUF model file. Instruct and chat models work best, from families like Llama and Mistral. To go back to a built-in model, click Remove.

Will it slow down my Mac?

It's built not to. Your keystrokes are handled first. After 15 minutes without typing, the model sleeps and frees its memory, then wakes the moment you type again. If your Mac runs short of memory, suggestions pause and come back on their own.

How quick it feels depends on your Mac and the model you pick.

Was this helpful?

If something’s missing, we’re a message away.

Related articles