Photo credit: www.theverge.com
OpenAI is poised to announce several new AI models in the near future, according to sources familiar with the developments reported by The Verge. One of the key releases anticipated is a reorganized iteration of the GPT-4 model, tentatively called GPT-4.1. This enhanced model is expected to improve upon the capabilities of the multi-modal GPT-4o, which integrates audio, visual, and textual data processing in real time.
Originally launched last year, GPT-4o has been a flagship offering. Reports suggest that along with the main GPT-4.1 model, OpenAI may also unveil smaller variants known as GPT-4.1 mini and nano, potentially as soon as next week.
In addition to GPT-4.1, OpenAI is preparing to roll out full versions of its o3 reasoning model and a mini version dubbed o4, which might be launched even sooner than the previously mentioned models. AI engineer Tibo Blaho identified references to these new models within an updated version of ChatGPT earlier today, indicating that the releases could happen imminently. Both the o3 and o4 mini editions are expected to go live next week, pending any changes to OpenAI’s current release schedule.
OpenAI CEO Sam Altman hinted on X about an exciting new feature set to launch today, though it remains uncertain whether this is connected to the announcements regarding the o3 and o4 mini models. However, it’s worth noting that OpenAI has faced delays in rolling out certain new models due to ongoing capacity constraints, which raises the possibility that the debut of the GPT-4.1 model could be postponed past its anticipated date next week. Any request for comments from OpenAI regarding this matter was not answered in time for publication.
Earlier this month, Altman noted on X that users might experience delays in new releases from OpenAI. He indicated that ongoing demand and system capacity issues might lead to slower performance and potential interruptions in service. This comes in light of increased pressure on the company’s infrastructure, particularly after the boost in demand for advanced image generation tools integrated into ChatGPT’s free tier, which Altman described as causing “our GPUs are melting.”
Source
www.theverge.com