tbh it's hard to say 'objectively' since a lot of it depends on what you're doing with it.
but like, if we're talking pure capabilities and widely accepted benchmarks, gpt-4 (especially 4o now) and claude 3 opus are usually at the top. gemini advanced (ultra) is also up there.
imo, you gotta try a few of them for your specific tasks cause they all have little quirks. for writing i might lean one way, for coding or logic problems another. they all excel in different areas really.
