This demo needs a Gemini API key with image generation access. Get one from aistudio.google.com. The key is saved in this browser's localStorage and passed directly to the Gemini client, without reloading or adding it to the URL.
Summon and Speak both use this key: Summon asks Gemini for the image, and Speak records what you say and sends it to Gemini to transcribe before summoning it.
Local prototyping only. Production apps must proxy AI calls through a
server you control. Legacy ?key= input is removed from
the address bar, but its initial request may already be in server
logs. Prefer entering your key here.