Show the AI an image (chat image / vision input)

You can give the AI an image so it can read it (describe, summarize, turn into a note, tag it, etc.). There are two ways.

[1] Attach an image directly:

• The 📎 button at the bottom-left of the input: pick an image from your Vault.
• Paste: paste a screenshot (or any image) straight into the input (⌘V).
• Drag & drop: drop image files from Finder / File Explorer onto the input.

Attached images appear as thumbnails above the input and can be removed with ×. Sent with your message, the AI sees them (up to 10 images per message).

[2] Refer to it by file name:

• Say something like "describe foo.png" or "look at the diagram and make a note" — the AI loads that Vault image (view_image) to see it. If the file is ambiguous, the AI looks it up from the list.

Supported formats: .png / .jpg / .jpeg / .gif / .webp. Images are downscaled to a 1568px long edge before sending. Images dropped/pasted from outside the Vault are saved into the Vault (the note folder, or the Vault root) before being attached.

Note: if the selected model does not support image input, you cannot send images (switch to a vision-capable model in Settings if you see an error).