In term of time and AD pathology, I would recommend the 3xTg, hardly used and you will find tons of papers about on pubmed. 5XFAD is good as well. I am using EFAD but most depend on what is your timeline and experimental design and whether or not you want to have a young vs old comparison.
If you are interested to study early phase of AD, I would recommend Tg2576 mouse model. Tg2576 mouse model recapitulates AD-related synaptic dysfunction and loss without neuronal death.
Based on what I know, there is no animal model that has all the features of AD so the best would be to have many animal models and the "total" of these models would be close to the definition of AD features.