OpenAI could launch its next big GPT AI model in December: What to expect

Reportedly, OpenAI will introduce an Orion AI model that is said to be up to 100 times more powerful than the GPT-4 model. It will be offered alongside he recently released GPT-4o model

Microsoft-backed artificial intelligence startup OpenAI is reportedly planning to launch its next big AI model by December. According to a report by The Verge, the new AI model, codenamed Orion internally, will be released by the end of this year.

 

The report stated that, unlike OpenAI’s current flagship models GPT-4o and the o1-series reasoning models, the anticipated Orion AI model will not initially be released on the ChatGPT chatbot. Instead, the new AI model will first be made available to partner companies, allowing them to build their own products and features. The report also mentioned that Microsoft, OpenAI’s main partner for deploying AI models, is already preparing to host the new AI model on its Azure cloud platform, which could be done as early as next month.

 

The Verge stated that an OpenAI executive informed the media outlet that the new Orion AI model is up to 100 times more powerful than the GPT-4 model and is separate from the newly released o1 model, which is focused entirely on reasoning and problem-solving. It is also reported that OpenAI has been using synthetic data provided by the o1 model to train Orion.

 

The report noted that OpenAI completed the training process for the new model last month, which coincided with CEO Sam Altman's post on X that hinted at the model's name. On September 14, Altman posted on X, stating, “excited for the winter constellations to rise soon; they are so great,” teasing the name Orion, which is also the name of a star constellation.

Internally, the company is said to believe that the new Orion model is a true successor to the GPT-4 model; however, it remains to be seen if OpenAI will name it GPT-5. Additionally, OpenAI could also push the release date to next year.