4o works just fine for me.
4o is way better at language stuff (e.g. writing)
o3-mini-high is way better at coding
Now I don't know what model to use for each task.
Is o4-mini-high better at coding than o3? How does o3 compare to o3-mini-high?
4o works just fine for me.
For me, 100%. I use it daily to connect to my xcode and do things I tell it to. Its like having a very fast junior developer and I can just focus on quality control.also plus, do you find it better / worth the money, I can't work it out for myself!
They replaced the o3-mini-high with o4-mini-high. So yes its an improved coding model, and in my test it just corrected an issue the o3 mini made (updated in the middle of a code application). It also seems much faster and is handling applies better (they also updated the chatgpt app so idk what fixed the janky code applications):4o is way better at language stuff (e.g. writing)
o3-mini-high is way better at coding
Now I don't know what model to use for each task.
Is o4-mini-high better at coding than o3? How does o3 compare to o3-mini-high?
I hope you’re right!Doesn't matter. Apple builds the hardware and operating systems that consumers use to access these competing services, so it's still a win for them whether Siri existed at all.
And Apple is rarely first to market. Doing it better is their goal, and the new Siri will be able to leverage the broader industry-wide models through Extensions support. Siri couldn't be in a better position to capitalize off what is happening in the market.
Mike Rockwell, good to have you in the forum here.Doesn't matter. Apple builds the hardware and operating systems that consumers use to access these competing services, so it's still a win for them whether Siri existed at all.
And Apple is rarely first to market. Doing it better is their goal, and the new Siri will be able to leverage the broader industry-wide models through Extensions support. Siri couldn't be in a better position to capitalize off what is happening in the market.
Agreed. Tried o4-mini-high for coding and it got tons of things wrong. Worse than o3-mini-high. Added slashes, apostrophes, and tags that didn’t belong. I’ve been using o3 to learn php and JavaScript and I fixed it all, then had 4o double check my work. The new models are just not that good. o3 never would have done that (never had any issues like o4 and it was persistent about creating errors even when called out)I'd put quotation marks around "smarter"