@thatlev: Sol 6.1 is getting an Ultrafast mode, and OpenAI says it can generate tokens up to eight times faster. It's announced for Codex in the days after launch, so the promise is much less time watching the model stream its answer. That doesn't make a whole coding task eight times faster, since thinking, tools and tests still take time. But when you're going back and forth on the same bug, faster output is a change you'll actually feel #OpenAI #Sol61 #Ultrafast #Codex #AIcoding
What's the point of speed? I don't need speed, I need quality of work
2026-10-03 11:10:23
20
vooxy :
make them kiss for free
2026-10-03 11:32:17
5
PanSage_ :
make them kiss
2026-10-03 19:17:21
1
Rktwo :
Guys instead of asking just go to my profile i have alot of kissing material *cough* i meant study
2026-10-04 00:18:41
0
Lok :
My question is will the reasoning effort is better, or it already good enough but OpenAI keep the generation speed slower in term of making user pay for it? Also is the output is equal with the normal speed. (Not an english native so sorry if my question write incorrectly)
2026-10-03 12:41:34
0
📀⌛️ :
say i wont
2026-10-03 11:06:57
9
RashE :
we don't need speed we need 10x efficiency I run out of tokens, I can run her all day if they let me
2026-10-03 13:07:23
0
аура картель :
у меня встал
2026-10-03 20:17:20
0
To see more videos from user @thatlev, please go to the Tikwm
homepage.