OpenAI is set to launch its new AI model, codenamed Orion, with deployment expected by December 2024. The launch marks another significant step forward for the tech giant, aiming to deliver a model that could be up to a hundred times more powerful than its predecessor, GPT-4. Unlike previous models such as GPT-4o and o1, which were widely released, Orion is aimed at a select group of small partners, focusing on specific applications and products.

Reliable sources indicate that engineers at Microsoft, OpenAI’s lead partner, are already preparing to integrate Orion into Azure by November, positioning the model for immediate use upon release. However, it remains uncertain whether OpenAI will officially classify Orion as GPT-5. The schedule for this release appears flexible, with a risk of delay.

The development of Orion is crucial as OpenAI strives to realize its long-term vision of artificial general intelligence (AGI). Recent efforts have included using the o1 model, Strawberry, to generate synthetic data for Orion’s formation, suggesting a robust preparation process for this ambitious model.

Hints of Orion’s release have been teased by OpenAI CEO Sam Altman, who recently posted about the “winter constellations” that would soon rise on social media. This subtle reference is thought to correlate with the name Orion, named after the winter constellation.

This upcoming launch is especially timely for OpenAI, which just completed a record organizational funding round, raising $6.6 billion. The company is currently facing a restructuring within its new for-profit model, alongside significant layoffs that have affected key staff, including CTO Mira Murati and CRO Bob McGrew.

As Orion prepares to launch, it promises to create major waves in the AI ​​industry, setting new standards for what artificial intelligence can achieve.