Which models are good for chat, research, reasoning (not coding)?
I recently decided to abandon my gemini subscription and get myself into APIs agregators like openrouter.ai and good open-source stuff like openWebUI and opencode. So a little offtop - I am using opencode + openchamber for agentic coding and this is amazing and not so hard to understand (with a little bit of digging through forums) which models are good/best for coding at this particular moment. So coding stuff is closed for me, I am happy with it. But the thing is I am not only coding with LLMs, I am brainstorming, researching, just asking about everyday stuff that I would google otherwise and etc, etc, etc. And I don't understand what models are good for this, seems like everyone’s talking only about coding related things.
I don't understand what should I pick for openWebUI. And seems that very little people are discussing that online comparing to amount of discussion on how good models at coding. To be honest I don't even understand what makes a model good for that type of tasks. What should I pay attention to? How good it at reasoning?