Text this: A Multi-Criteria Comparison of Large Language Model Powered Assistants in Pre-Research Studies for the Academia