¿Gemini o Chatgpt?

Overall, I've always really liked chatgpt, and I've had good workflows with it (in fact, I'm currently paying for the subscription). I use it mainly to help me reason ideas (because although it is usually very positivist, with the correct command it gives good critical points of view) and to help me study (not so much general information but rather very specific questions that are NOT SO important as to deserve a dedication of searching but that I feel help me to better dimension a topic), and this has the advantage of having an important mode for my career which is the “PubMed Boddy”, which claims to only use data from pubmed, a very valid medical source. Now, lately YouTube has been bombarding me with videos talking about the new Gemini, and how it is superior to chatgpt. The problem is that they generally talk about superiority in functions for which I do not use it (such as writing texts, solving mathematical problems or editing images), and sometimes I have used another Google tool (Notebooke LM) to help me have more general ideas on very extensive topics but I feel that it always ends up losing important information from the sources that I give it.
And that's when I wonder, how can I really know which AI is best for such specific things? Sure, using both and seeing which one I like better would sound like a direct and easy idea, but I'm afraid I'll end up choosing aesthetics over quality, for example, that I like the way it explains information better but it's actually less accurate or complete, or that I like some people's opinions more but unconsciously it's just because they're what I want to read. I don't know if I understood correctly, does anyone have any ideas?

PS: Someone might be thinking about open evidence as an alternative for direct questions, and yes, it is a good alternative, but it has 2 problems:
1. It falls too much, which makes it unreliable in times of important doubt.
2. I feel like the answer is very to the point, so I don't get the context right.