FAQ
Quick answers to common questions about RAZE.
Installation
Install raze-cli with pipx install raze-cli, then run raze onboard. That's it.
You need Python available to install raze-cli, but the product keeps that requirement simple and practical. You don't need to be a Python developer.
Not for normal setup. Some advanced Docker-backed employee workflows may require it, but the main dashboard and local workflow work without Docker.
Updates
Update the CLI with pip install --upgrade raze-cli.
Use raze update to update the managed runtime bundle.
The managed runtime bundle updates the dashboard and runtime. The CLI and dashboard are on separate update channels.
Providers & Models
Yes. RAZE can work with local model setups where supported, such as local-provider flows.
Yes. RAZE supports provider-based setups as part of the product flow.
RAZE can still open the dashboard and guide setup. The dashboard will help you connect a provider or local model profile.
Use local models when you want more control or privacy. Use API providers when you want easier access to hosted models.
Data & Privacy
RAZE is local-first. Your runtime, dashboard state, and configuration live on your machine. If you connect third-party providers, requests you send to those providers are subject to those providers' terms.
RAZE stores local runtime state, configuration, logs, and related data on your machine.
Because the product is designed to run locally, keep runtime control in your hands, and let you choose your own providers and models.
Requirements
A machine that can run the RAZE runtime, Python to install raze-cli, a first run with raze onboard, and at least one configured provider or local model.
Docker-backed workflows, manual bundle paths, manifest overrides, advanced provider-profile tuning, and custom local model setups.