Using a Smart Filter with “Instance is Duplicate” I get a list of 20 odd files that are duplicates in my database. But how do i delete the duplicates quickly?
In the Smart filter view the Delete icon is greyed out.
I can click on Reveal in the Data menu but that doesn’t actually reveal the file.
Is there an easy way to delete one instance of each duplicate or at least to delete them one by one within the smart filter window?
Just use Data > Move All Instances To Trash. Or the shortcut Command-Option-Backspace. And the next beta will fix the script to easily remove duplicates.
Thanks Christian.
Am I misunderstanding something - surely one would normally want to send all but one instance to Trash because you want to keep one instance but remove the other duplicates?
Yes, that is exactly what I want: choose one instance and make DT delete all (other) duplicats. And the same with replicants.
Is there - or will be - a script for that?
Or an implementation in the next beta? 
Have you looked at the application’s Script menu > Data > Move Duplicates To Trash?
It seems that this is exactly what I want. Thanks, Annard. I had a look at it but the prompt »Kill in 1 selected items:« was too confusing for me. So I didn’t try that script.
Unfortunately in the actual beta state I may not use it because dtpo has quite a misconception of what a duplicate is, right?
But I’m sure that will be repaired in the near future.
I don’t think so. What exactly does not work as expected?
In the »duplicate« smart group I found some files with nearly the same date, nearly the same size, nearly the same name and nearly the same content but different documents like e.g. invoices.
It wouldn’t be clever to choose one and say »Kill the rest«
Maybe what Greg_Jones reported in:
Duplicates that are not actual duplicates
I’ve noticed that sort of seemingly overzealous duplicate detection here, too.
DEVONthink is only comparing the contents, therefore things like name, date or size don’t matter. Do these invoices contain text? Is the text identical?
The are »pdf+text«-files and, yes, the content is different.
E. g. google-adwords invoices with different dates and different amounts.
Please try this:
Are these texts still marked as duplicates?
I did.
The two text-files in the new DB aren’t duplicates anymore.
The two pdf + teft files in the new DB aren’t duplicates anymore.
But the text-file and the concerning pdf + text file are duplicats. What imo isn’t quite correct.
Is the solution to hustle all the stuff into a new db?
The text of both files is identical, therefore this is at least somewhat correct. However, the next beta will improve the handling of duplicates but it will be never an exact comparison of files.
»isn’t quite correct« and » is at least somewhat correct« means the same, right? 
Maybe 
I too am having trouble with this. I select and run the script to remove duplicates, but nothing happens and the script doesn’t prompt me. I tried selecting a group of files and just one instance of the file hoping it would remove things, but nothing happens. I have the scripts installed… How exactly does this work?
The script does not work due to a bug of the latest beta.