pandas str.extract multiple regex groups with OR

[pythonidae]

Hi. I have a column of text with inputs like, 3" deep, 4 inches deep, 5" depth. I wanted to extract just the value and put this in a new column.

From regex101.com, I confirm the following captures any of these for all my examples:

r'(\d)\"\s*deep|(\d)\"\s*depth|(\d)\sinches\sdeep'

However, when I try to use this regex pattern with str.extract as follows

df['depth']=df['text'].str.extract(r'(\d)\"\s*deep|(\d)\"\s*depth|(\d)\sinches\sdeep')

Error:
ValueError: Wrong number of items passed 3, placement implies 1

The docs say it should just return the first match. I tried str.extractall just in case, but this produces

Error:
TypeError: incompatible index of inserted column with frame index

How do I pass regex OR patterns to str.extract? Or, how should I match each of these cases to create a new column?

Thanks!