OpenAI is preparing to launch the advanced Orion model on this date.

OpenAI, a leading artificial intelligence company, is preparing to announce its new advanced model codenamed “Orion,” expected to be released by December. According to a recent report by The Verge, OpenAI will take a different approach in unveiling this model compared to previous versions like GPT-4o and o1.
Instead of launching “Orion” widely on the ChatGPT platform, OpenAI plans to give partner companies initial access to the new model, enabling them to build their products and features relying on it. This move is part of OpenAI’s strategy to expand the use of its models in different business environments.
In a related context, sources have indicated that Microsoft engineers are working to host the “Orion” model on the Azure platform for cloud computing by November. “Orion” is believed to represent the next generation of the GPT-4 model, although it’s unclear whether the company will name it “GPT-5” upon release.
Both OpenAI and Microsoft have declined to comment on this information, while an official spokesperson from OpenAI told TechCrunch that the report was “inaccurate” without providing further details.
The “Orion” model is expected to be up to a hundred times more powerful than the current GPT-4 model, according to statements from OpenAI officials. Researchers at the company celebrated completing the training of the new model during an informal event.
The new model stands out as it is separate from the o1 thinking and reasoning model released by OpenAI in September, with the company aiming to integrate Large Language Models (LLMs) in the long term to develop a more powerful model, potentially dubbed as “Artificial General Intelligence AGI,” capable of simulating human intelligence and outperforming it.
This comes at a critical time for OpenAI, as they recently concluded a historic funding round of up to $6 billion, requiring them to restructure to become a profit-driven entity to meet investor expectations for continuously releasing advanced models to maintain leadership in the field of artificial intelligence.