OpenAI may launch new GPT model December

OpenAI, backed by Microsoft, is expected to launch its next major AI model by December, according to The Verge.
This new model will differ from OpenAI’s current flagship models, GPT-4o and the o1-series, as it won’t initially be available on the ChatGPT chatbot. Instead, Orion will first be offered to partner companies for developing their own products. Microsoft, OpenAI’s primary partner, is already preparing to host the model on its Azure cloud platform, potentially as soon as next month. An OpenAI executive revealed that Orion is up to 100 times more powerful than GPT-4, and unlike the o1 model, which focuses on reasoning, Orion is being trained with synthetic data from the o1 model.
Training for Orion was completed last month, coinciding with CEO Sam Altman’s cryptic post on X (formerly Twitter), where he hinted at the model’s name. On September 14, Altman teased, “excited for the winter constellations to rise soon; they are so great,” referring to Orion, the star constellation. While the company internally sees Orion as a true successor to GPT-4, it’s unclear if it will be called GPT-5. OpenAI may even delay its release until next year.