Re:OpenAI O3-Mini
by siliconc0w on Friday January 31, 2025 @02:20PM
(#42890773)
The real heated contest here amongst the top AI labs is to see who can come up with the most confusing product names.
[ Parent
| on HN ]
It's nice to see Google finally having competition in a space it used to really dominate (though they definitely still are holding their own with all the Gemini naming). I feel like it takes real effort to have product names be this confusing and capricious
Gemini naming seems pretty straightforward at this point. 2.0 is the full model, flash is a smaller/faster/cheaper model, and flash thinking is a smaller/faster/cheaper reasoning model with Cost.
> 2.0 is the full model
Not quite. "2.0 Flash" is also called 2.0. The "Pro" models are the full models. But, I love how they have both "gemini-exp-1206" and "gemini-2.0-flash-thinking-exp-01-21". The first one doesn't even say what type of model it is, presumably it should have been "gemini-2.0-pro-exp-1206", but they didn't want to label it that for some reason, and now they're putting a hyphen in the date string where they weren't before.
Not to mention they have both "Flash" and "Flash-8B"... which I think will confuse people. IMO, it should be "Flash-${Parameters}B" for both of them if they're going to mention it for one.
But, I generally think Google's Gemini naming structure has been pretty decent.
Re:OpenAI O3-Mini
by not_a_bot_4sho on Friday January 31, 2025 @02:39PM
(#42891018)
Someone dropped the ball with Phi models. There is clearly an opportunity for XP and Ultimate and X/S editions.
Personally waiting for the ME model. Should be great at jokes and humor.
I really think a "OpenAI Me" is what's needed.
Re:OpenAI O3-Mini
by TheOtherHobbes on Friday January 31, 2025 @03:10PM
(#42891347)
Surprised Apple hasn't gone with iI Pro Max.