OpenAI is getting ready to launch as many as three new AI fashions, presumably referred to as “o4-mini”, “o4-mini-high” and “o3”.
Proper now, ChatGPT has as many as 5 fashions, together with GPT 4o (the non-reasoning mannequin), GPT 4.5 (one other non-reasoning mannequin however with larger creativity), and three reasoning fashions: o1, o3-mini, and o3-mini-high.
o1’s successor is o3, however a full-fledged mannequin is not obtainable but.
We solely have entry to the o3-mini and o3-mini-high, that are small reasoning fashions within the o-series, providing improved efficiency and quicker responses.
Supply: BleepingComputer
In accordance with references discovered on ChatGPT’s internet app, OpenAI is getting ready to launch as many as three new fashions: o3 (a full-fledged reasoning mannequin within the o3-mini, o3-mini-high lineup), and o4-mini, o4-mini-high, just like the present fashions however with larger reasoning capabilities.
In a publish on X final week, OpenAI CEO Sam Altman beforehand confirmed that OpenAI plans to launch new o3 and o4 fashions earlier than the upcoming GPT-5, which is alleged to be a unified reasoning mannequin.
“We’re going to launch o3 and o4-mini in spite of everything, in all probability in a few weeks, after which do GPT-5 in a couple of months,” Sam Altman wrote in a publish on X.
“There are a bunch of causes for this, however probably the most thrilling one is that we’re going to have the ability to make GPT-5 significantly better than we initially thought. We additionally discovered it tougher than we thought it was going to be to easily combine all the pieces, and we need to ensure that now we have sufficient capability to help what we count on to be unprecedented demand.”
We do not know when OpenAI plans to ship these three new fashions, however preparations are clearly underway primarily based on the references.
Based mostly on an evaluation of 14M malicious actions, uncover the highest 10 MITRE ATT&CK strategies behind 93% of assaults and easy methods to defend towards them.
