User level: Level 1 (73 points)When I try to use the "use model" command in Shortcuts, the documentation says I should be able to pass it both a prompt and a file to analyze. But there doesn't appear to be any way to give it a prompt. I am using ChatGPT, which is linked to my paid ChatGPT subscription.
also, I am passing it an audio file, but it is somehow reading it as a bunch of metadata about a video.
Here's what it looks like:
help?
[Edited by Moderator]
Posted on Oct 3, 2026 4:08 PM
User level: Level 7 (31,566 points)Oct 4, 2026 2:20 AM in response to Nospam Nospam
Although hat 'Record audio' action stores the audio file in the Voice Memos app, it does not make it available to subsequent actions in the shortcut. Only metadata about the audio file is returned — see a related situation here:
Save Voice Memos via Shortcuts to iCloud/Files?
Feedback - Voice Memos - Apple
õ¿õ¬
User level: Level 1 (73 points)Oct 4, 2026 3:07 AM in response to Alancito
Thanks for the explanation! Frustrating, but at least I know I can stop wasting my time trying to make it work.
Any idea how the ChatGPT "analyze image" shortcut works? It seems to pass an image and a prompt to ChatGPT; feels like the same thing I want to do except that it works with images and not with sounds.
Originally posted at discussions.apple.com/thread/256369937