This has been my experience as well. Grok will end tasks almost immediately and claim "Done!". It's definitely the laziest and most "dishonest" of all the model

This has been my experience as well. Grok will end tasks almost immediately and claim "Done!". It's definitely the laziest and most "dishonest" of all the models. The others aren't perfect, but I can't use Grok for any serious coding task.