start here
install EV in three steps.
Get the desktop assistant running, connect a provider, and make your first voice request. The app uses your own keys, so you stay in control of your usage.
02
give EV permission to help
On first launch, allow microphone access so EV can hear you. If you want screen awareness or computer control, enable screen recording and accessibility permissions in your operating system settings.
microphone
screen recording
accessibility
Keep computer control supervised. AI can make mistakes, especially around irreversible actions.
03
connect your first API key
Open EV settings and paste one provider key. Start with OpenAI for the smoothest default experience, or use the provider roadmap in the guide to choose another model stack.
# example local environment values
OPENAI_API_KEY=your_key_here
# never commit keys to git or share them in screenshotsAfter saving, say “hey EV” and ask a simple question. If the key is rejected, check the provider dashboard, account credits, and whether the selected model is available in your region.