What you need
You need a Mac with Apple silicon and macOS 14 or later. ASIST does not run on an Intel Mac.
Speech recognition and the other local models run on your Mac. With all of them in use, they take about 4.1 GB of memory in total, and choosing Qwen3-TTS for reading aloud takes about 2 GB more. You can choose Qwen3-TTS on a Mac with 16 GB of memory or more. The breakdown is in Memory requirements.
An API key for the conversation model
Section titled “An API key for the conversation model”You need an API key from one of these providers. You enter it during the first-time setup. Usage is billed to your account with that provider.
| Provider | Where to create a key |
|---|---|
| Anthropic | console.anthropic.com |
| OpenAI | platform.openai.com |
| aistudio.google.com | |
| Cerebras | cloud.cerebras.ai |
Optional
Section titled “Optional”| What | What it is for |
|---|---|
The codex or claude CLI |
Agent jobs and memory curation. Install the one you want to use and sign in to it first. See The Agent CLI. |
| VOICEVOX or AivisSpeech | Reading Japanese aloud. If it is in your Applications folder, ASIST starts it in the background and connects to it. |
The uv bundled with the app creates the Python environments for the models that run on your Mac. uv also downloads Python at that point, so you do not need to install anything beforehand.
Before you start
Section titled “Before you start”ASIST’s answers can be wrong, the APIs are billed to your own accounts, and an Agent job you approve can change your files. Before you start, read Using ASIST safely.