OpenAI’s CEO, Sam Altman, has announced delaying the launch of the firm’s latest base model, the GPT-5, after facing demand for its recently launched native image generation feature. He stated that o3 and 04-mini will be launched in a couple of weeks.
Notably, ChatGPT’s underlying model, GPT-4o, received a massive update last week that allowed it to directly generate real picture images into Japanese animated studio Ghibli art. The new features allowed ChatGPT to better understand the user’s context while generating more advanced and accurate images. As a result of its advanced features, it created a giant trend on social media globally.
Sam Altman has declared this on his X post, stating that, “Change of plans: we are going to release o3 and o4-mini after all, probably in a couple of weeks, and then do GPT-5 in a few months.”
Altman has not stated specific reasons behind it, he just stated that there are several reasons for delaying GPT-5, but the best one is that this extra time will let them make GPT-5 much better than they first planned.
On 15th February Altman has explain the roadmap for GPT-5 on his X post stating that, “A top goal for us is to unify o-series models and GPT-series models by creating systems that can use all our tools, know when to think for a long time or not, and generally be useful for an expansive range of tasks.”
As per the reports, OpenAI’s newer “o” series models, like o3 Mini and o1, are designed to think more like humans. They focus on reasoning to tackle tough tasks like coding, math, and problem-solving. These models are more detailed but slower than regular models, so they’re not the best for simple questions.
GPT-5 will combine features from their current models into one. Everyone will get “unlimited access” to GPT-5 at a normal intelligence level, but people who pay for Plus or Pro plans will get an even smarter version.
OpenAI first talked about its o3 and o3 Mini models during a 12-day event in December. This was due to the introduction of free reasoning features by a Chinese company, DeepSeek AI.
OpenAI quickly released o3 Mini in late January and let free users try it with some limits. Now, for the first time since then, OpenAI has hinted at when they might fully release o3.
Besides these, OpenAI is also planning to launch a new “open weights” model, the first since GPT-2, in the following months. A part of this, on 2nd April, Sam Altman has also hinted at the launch of the firm’s Image Version 2.
Also Read: After Ghibli success, Altman Hints At OpenAI Image Version 2.0
