AI Assistance

The editor’s footer has an AI input with a provider picker. Two backends:

• Apple Intelligence — on-device (macOS 26+, Apple silicon). No API key.
• Google Gemini — cloud; requires an API key (Settings → AI providers).

What you can ask

• Questions about the image — “what does the highlighted text say?”
• Draw annotations — “circle the total”, “box every button”.
• Edit the image — “remove the person in the background”, “make it 16:9” (Gemini image editing).

The picker’s tips button (ⓘ) lists example prompts for the active provider. Press ↑ / ↓ in the input to recall previous prompts.