Very curious for what use cases you're finding gemini unusable.

Scientific research and proof-reading. Gemini is the laziest LLM I've used. Frequently he will lie that he searched for something and just make stuff up, basically never happens to me when I'm using gpt5.2.

▲

flexagoon an hour ago | parent [-]

Do you use it directly? I've only used it though Kagi Assistant but it works better than any other model for me

	▲	azan_ an hour ago \| parent [-]
		Yes, only directly (I mean through the default gemini interface, not API).

▲

double0jimb0 2 hours ago | parent | prev [-]

In my experience with Gemini, I find it incapable of not hallucinating.