OpenAI Delays GPT-5, Confirms o3 Release

After pulling the plug on the consumer launch of its o3 reasoning model back in February, OpenAI is now gearing up to drop both o3 and its next-gen counterpart, o4-mini, in just "a couple of weeks." This news came straight from the horse's mouth, with OpenAI's CEO Sam Altman spilling the beans on X last Friday.
Altman let slip that this change of heart is all tied up with OpenAI's upcoming biggie, GPT-5. They've been pretty hush-hush about it, but word on the street is that GPT-5 is gonna be a one-stop shop, rolling reasoning capabilities into the mix.
"We're gonna make GPT-5 way better than we first thought," Altman scribbled in his post. "But, man, it's been tougher than we expected to get everything to play nice together. And we gotta make sure we've got the juice to handle what we think is gonna be a crazy amount of demand."
He also dropped a hint that we might see GPT-5 hit the scene "in a few months" — a bit later than what they'd hoped for at first.
From what OpenAI's let slip about GPT-5, they're planning to let everyone chat with it at the "standard intelligence setting," but don't go overboard, 'cause there are limits to stop any abuse. If you're a ChatGPT Plus member, you'll get to crank it up a notch to a "higher level of intelligence," and if you're on the ChatGPT Pro plan, you'll be able to push it to an "even higher level of intelligence."
Back in the early days of the year, Altman was all about how GPT-5 would be the Swiss Army knife of AI, packing in voice, Canvas, search, deep research, and more. "We're aiming to bring all our models together," he said, "so they can use all our tools, know when to take their time, and be super handy for a bunch of different jobs."
OpenAI's feeling the heat from rivals like the Chinese AI lab DeepSeek, who are playing it fast and loose with an "open" approach to launching their models. Unlike OpenAI's more guarded strategy, these "open" folks are letting the AI community tinker with and even make a buck off their models.
Alongside o3, o3 pro, o4-mini, and GPT-5, OpenAI's also planning to roll out its first open language model since GPT-2 in the coming months. Altman mentioned earlier this week that this new kid on the block will have reasoning chops and will go through some extra safety checks before hitting the streets.
Related article
OpenAI Commits to Fixes After ChatGPT's Overly Agreeable Responses
OpenAI plans to revise its AI model update process for ChatGPT after an update caused excessively sycophantic responses, prompting widespread user feedback.Last weekend, following an update to GPT-4o,
OpenAI Acquires Talent from AI Recommendation Startup Crossing Minds
Crossing Minds, a startup specializing in AI-driven recommendation systems for e-commerce, announced Thursday that its team is transitioning to OpenAI.Supported by investors including Index Ventures,
OpenAI to Build Massive 5-Gigawatt AI Data Center in Abu Dhabi
OpenAI is set to spearhead a colossal 5-gigawatt data center campus in Abu Dhabi, establishing itself as a key tenant in one of the world’s largest AI infrastructure initiatives, per a recent Bloomber
Comments (29)
0/200
KeithMoore
August 1, 2025 at 2:45:46 AM EDT
Super curious about o3's release! 😎 Will it outsmart us all or just be another hyped-up model? Can't wait to test it!
0
RogerGonzalez
July 31, 2025 at 10:55:47 PM EDT
Super hyped for o3 and o4-mini dropping soon! 😎 OpenAI's been cooking something big, but I’m curious if it’ll outshine competitors or just add to the AI hype train.
0
RonaldMitchell
July 31, 2025 at 10:48:18 PM EDT
Sounds like OpenAI's playing a wild card with o3 and o4-mini! Can't wait to see how these models stack up. 😎
0
HarperGreen
July 27, 2025 at 9:19:05 PM EDT
C'est fou comme OpenAI joue avec nos nerfs ! 😅 Annuler o3, puis le relancer avec o4-mini si vite... Ça sent la course contre la montre avec xAI. Vous pensez qu’ils vont vraiment révolutionner l’IA ou juste faire du buzz ?
0
StephenGreen
April 22, 2025 at 12:32:30 AM EDT
GPT-5の遅延は残念だけど、o3が楽しみ!どんなことができるのか早く見たいな。o4-miniも興味深いね。ただ、待たされるのが長引かないことを願ってるよ。スムーズなローンチを祈ってる!🙏
0
CharlesJohnson
April 21, 2025 at 8:26:19 AM EDT
La demora de OpenAI con GPT-5 es una decepción, pero estoy emocionado por o3. No puedo esperar para ver lo que puede hacer. El o4-mini también suena intrigante. Solo espero que no nos hagan esperar demasiado. ¡Cruzo los dedos por un lanzamiento sin problemas! 🤞
0
After pulling the plug on the consumer launch of its o3 reasoning model back in February, OpenAI is now gearing up to drop both o3 and its next-gen counterpart, o4-mini, in just "a couple of weeks." This news came straight from the horse's mouth, with OpenAI's CEO Sam Altman spilling the beans on X last Friday.
Altman let slip that this change of heart is all tied up with OpenAI's upcoming biggie, GPT-5. They've been pretty hush-hush about it, but word on the street is that GPT-5 is gonna be a one-stop shop, rolling reasoning capabilities into the mix.
"We're gonna make GPT-5 way better than we first thought," Altman scribbled in his post. "But, man, it's been tougher than we expected to get everything to play nice together. And we gotta make sure we've got the juice to handle what we think is gonna be a crazy amount of demand."
He also dropped a hint that we might see GPT-5 hit the scene "in a few months" — a bit later than what they'd hoped for at first.
From what OpenAI's let slip about GPT-5, they're planning to let everyone chat with it at the "standard intelligence setting," but don't go overboard, 'cause there are limits to stop any abuse. If you're a ChatGPT Plus member, you'll get to crank it up a notch to a "higher level of intelligence," and if you're on the ChatGPT Pro plan, you'll be able to push it to an "even higher level of intelligence."
Back in the early days of the year, Altman was all about how GPT-5 would be the Swiss Army knife of AI, packing in voice, Canvas, search, deep research, and more. "We're aiming to bring all our models together," he said, "so they can use all our tools, know when to take their time, and be super handy for a bunch of different jobs."
OpenAI's feeling the heat from rivals like the Chinese AI lab DeepSeek, who are playing it fast and loose with an "open" approach to launching their models. Unlike OpenAI's more guarded strategy, these "open" folks are letting the AI community tinker with and even make a buck off their models.
Alongside o3, o3 pro, o4-mini, and GPT-5, OpenAI's also planning to roll out its first open language model since GPT-2 in the coming months. Altman mentioned earlier this week that this new kid on the block will have reasoning chops and will go through some extra safety checks before hitting the streets.




Super curious about o3's release! 😎 Will it outsmart us all or just be another hyped-up model? Can't wait to test it!




Super hyped for o3 and o4-mini dropping soon! 😎 OpenAI's been cooking something big, but I’m curious if it’ll outshine competitors or just add to the AI hype train.




Sounds like OpenAI's playing a wild card with o3 and o4-mini! Can't wait to see how these models stack up. 😎




C'est fou comme OpenAI joue avec nos nerfs ! 😅 Annuler o3, puis le relancer avec o4-mini si vite... Ça sent la course contre la montre avec xAI. Vous pensez qu’ils vont vraiment révolutionner l’IA ou juste faire du buzz ?




GPT-5の遅延は残念だけど、o3が楽しみ!どんなことができるのか早く見たいな。o4-miniも興味深いね。ただ、待たされるのが長引かないことを願ってるよ。スムーズなローンチを祈ってる!🙏




La demora de OpenAI con GPT-5 es una decepción, pero estoy emocionado por o3. No puedo esperar para ver lo que puede hacer. El o4-mini también suena intrigante. Solo espero que no nos hagan esperar demasiado. ¡Cruzo los dedos por un lanzamiento sin problemas! 🤞












