OpenAI is reportedly developing a new AI model called GPT-4.1, an enhanced version of its multimodal model GPT-4o.
This news comes amid growing anticipation for the next big leap in generative AI, while GPT-5 still appears to be a distant milestone.
According to a recent report by The Verge, OpenAI is already testing GPT-4.1, along with other model variants such as o3, o4-mini, and GPT-4.1 mini and nano, on its API platform. These findings were first spotted by AI researcher Tibor Blaho, further fueling speculation about GPT-4.1’s imminent release.
Unlike GPT-4.5, which reportedly focuses more on creativity and response quality, GPT-4.1 is believed to be a direct successor to GPT-4o, OpenAI’s real-time, multimodal model. While GPT-4.5 was teased earlier in a training video, OpenAI CEO Sam Altman hinted that the company is exploring ways to retrain GPT-4 from scratch using improved data and systems.
OpenAI seems to prioritize updates to its existing infrastructure in the current development cycle over a full leap to GPT-5. With GPT-4.1 and its scaled-down variants already in testing, users can likely expect enhanced performance and broader accessibility across devices and use cases soon.
