CEO Sam Altman believes that the way OpenAI names its current models is confusing and promises to fix it in the next few months.
"Why don't we fix the names of our AI models this summer?" Altman asked on X on April 15. "Until then, people will have a few more months to make fun of us, which we absolutely deserve."
According to Bussiness Insider, OpenAI's naming is confusing to users and makes it difficult to remember and distinguish the differences between models, for example GPT-4.1, GPT-4.1 mini, GPT-4.1 nano, GPT-4o, GPT-4o mini, o1-pro, o3-mini.
In early April, OpenAI product manager Kevin Weil also commented: "The naming is terrible. We'll address it at some point. It's not the most important thing, so we haven't spent much time on it. We'll come up with a new name, like 'o3 mini high. '"
Altman has said he hates the name ChatGPT. "It's a terrible name, but it's probably so ubiquitous that it's never going to change," he said on comedian Trevor Noah's What Now? Podcast in 2023.
On April 14, OpenAI announced three new AI models: the standard GPT-4.1, the smaller and cheaper GPT-4.1 mini, and the company's "smallest, fastest, and cheapest" GPT-4.1 nano. The company said GPT-4.1 outperforms GPT-4o "in almost every way," notably its ability to program and follow instructions. The new model can also process up to one million context tokens (text, images, videos) in a single command, much more than the 128,000 tokens limit of GPT-4o.
GPT-4.1 is also 26% cheaper than its predecessor. TechCrunch rates this as a solution to compete with low-cost models like DeepSeek from China.
(via Business Insider, TechCrunch)