ECNETNews reports that the CEO of OpenAI, Sam Altman, has revealed that the highly anticipated GPT-4.5 will “likely” start to be rolled out to ChatGPT Plus subscribers over the span of a few days, rather than all at once. Users are encouraged to remain patient during this process.
In a statement, Altman noted, “there is no perfect way to do this; we wanted to do it for everyone tomorrow, but it would have meant launching with a very low rate limit. We believe people are going to use this a lot and love it, so it’s better to allow users to engage in meaningful conversations without overwhelming the system.”
Introduced last Thursday, GPT-4.5 represents the latest evolution of OpenAI’s large language model (LLM) and is expected to be the final version before the implementation of chain-of-thought prompting. This technique aims to emulate human reasoning by breaking down complex tasks into logical, manageable steps.
While GPT-4.5 has not fully adopted this approach yet, early testing suggests that interactions with it “feel more natural,” according to OpenAI.
OpenAI stated, “Its broader knowledge base, improved ability to follow user intent, and greater emotional intelligence make it useful for tasks like enhancing writing, programming, and solving practical problems. We also expect it to hallucinate less.”
Despite the enhancements, Altman tempered expectations by clarifying that GPT-4.5 “isn’t a reasoning model and won’t crush benchmarks.”
The staggered rollout of GPT-4.5 aligns with previous comments from Altman, indicating that technical limitations related to graphics processing units (GPUs) initially hindered a simultaneous launch for ChatGPT Plus and Pro subscribers. Currently, ChatGPT Pro subscribers are the first to experience the new model, with a subscription priced at $200 per month, while ChatGPT Plus is available for $20 per month.
“We will be adding tens of thousands of GPUs next week and rolling it out to the Plus tier then,” Altman wrote. “Hundreds of thousands are coming soon, and we anticipate that users will utilize every resource we can secure.”
