Microsoft (MS) has decided to supply OpenAI's 'GPT-4 Turbo', which supports the paid version of CoPilot, to the free version. In light of this, evaluation has emerged that the launch of a latest ‘GPT-4.5’ model is imminent. The reason is that not only MS but in addition Open AI needs a latest, cutting-edge model.

CNET reported on the 14th (local time) that Microsoft upgraded Co-Pilot in order that even free Co-Pilot users can use GPT-4 Turbo. Just set CoPilot to Creative or Precise mode.

GPT-4 Turbo has been trained on data as much as April 2023 and supports a context window of 128,000 tokens. It also provides a text-to-voice conversion function and a 'Dali 3'-based image creation function.

Co-Pilot Pro subscribers may also use OpenAI's latest model, GPT-4 Turbo, in addition to 'Co-Pilot GPT Builder', a customized chatbot tool released on the eleventh. Nevertheless, even when you pay $20 per 30 days, there isn’t any differentiation when you use the identical model as free users.

Particularly, MS is working hard to expand the Co-Pilot paid model. On at the present time, Co-Pilot Pro subscriptions were expanded to 222 countries world wide.

To draw more users to the paid service, the corporate may also launch a one-month free trial of CoPilot Pro. Mobile also announced that it would implement a free trial inside just a few months.

Due to this case, it’s predicted that Co-Pilot Pro will soon be equipped with a model more powerful than the prevailing GPT-4 Turbo.

OpenAI, which had been silent about GPT-4.5, now has a reason to release it. It is because Antropic's 'Claude 3', which appeared last week, pushed out GPT-4 and took the title of 'currently strongest'.

It's not only a matter of pride. As a way to avoid falling behind Claude 3 in enterprise services, a more powerful model have to be released.

Because it is simply too early to release GPT-5, the main focus is of course on GPT-4.5. Rumors have been circulating in most communities since last week that OpenAI will announce GPT-4.5 or GPT-5.

Here we hear that a notice about 'GPT-4.5 Turbo' has been leaked, and it now looks as if a fait accompli.

A Reddit user claimed that earlier this week OpenAI mistakenly published a blog post that said “GPT-4.5 Turbo outperforms GPT-4 Turbo in speed, accuracy, and scalability.”

The blog post was quickly deleted from the Open AI website.

Reporter Park Chan cpark@aitimes.com