That appears to be working although there isn’t any data that looks like it would cause a problem, I have removed the slice from the DeepL node as well just to do it all.
Hmm strange because I was having this error on each of my DeepL workflows, independent of the RSS feed used to feed the translation, I know this one has longer text, maybe you can reproduce it with this one Russia-State via CyberBenB on Inoreader
I had to limit the slice at 200 to be able to run a translation without error.
In my current production stack (cloudron.io) I will have to wait for the release to be done on your side for the cloudron team to upgrade their package of N8N, usually it goes fast, let’s hope for the merge !