OpenAI, supported by Microsoft, is reportedly gearing up to release its latest AI model, codenamed Orion, by the end of the year. As per The Verge, Orion’s launch marks a significant advancement from OpenAI’s current flagship models, such as GPT-4 and the o1-series reasoning model.
Unlike previous releases, Orion will initially be made available to OpenAI's partner companies, who can use it to build custom products and features. Microsoft, a primary partner, is preparing to host Orion on its Azure cloud platform as early as next month.
An OpenAI executive shared with The Verge that Orion is up to 100 times more powerful than GPT-4 and has been trained using synthetic data generated by the o1 reasoning model. OpenAI completed Orion's training last month, around the time CEO Sam Altman hinted at the model’s name with a post on X: “excited for the winter constellations to rise soon; they are so great.” This appears to reference Orion, the prominent star constellation.
Comments