TechFlow reports, on October 25, according to The Verge, OpenAI plans to launch its next-generation AI model codenamed Orion in December 2024. Unlike GPT-4o and o1, Orion will not be widely released through ChatGPT initially; instead, it will be prioritized for licensing to enterprises closely collaborating with OpenAI to develop their own products and features.
Insiders reveal that Microsoft engineers are preparing to deploy Orion on Azure as early as November. While OpenAI internally views Orion as the successor to GPT-4, it remains unclear whether it will be publicly branded as GPT-5. A senior OpenAI executive previously suggested that Orion could be up to 100 times more powerful than GPT-4.
In September, OpenAI CEO Sam Altman posted on social media about winter constellations, which was interpreted as hinting at Orion's development progress. That same month, company researchers held a celebration marking the completion of training the new model.
