GPT-5 set to be launched by December says The Verge

OpenAI, the artificial intelligence startup supported by Microsoft, is reportedly preparing to launch its next significant AI model GPT-5 by December. According to a detailed report by The Verge, this new system, which has been internally codenamed "Orion," is expected to be rolled out before the end of the year, making it one of the most anticipated AI releases of the season.

This new AI model will differ from OpenAI’s flagship offerings, such as GPT-4 and the recently released O1-series reasoning models. Unlike those models, which are available to the public through OpenAI’s widely used ChatGPT platform, the Orion model will not be accessible to users through ChatGPT initially. Instead, OpenAI plans to partner with various companies, granting them exclusive access to the model so they can use it to build their own custom products and AI features. This move could pave the way for new AI applications across industries by allowing partner companies to leverage the power of the Orion model to enhance their offerings.

Moreover, Microsoft, OpenAI’s primary partner for deploying and scaling AI solutions, is preparing to host the Orion model on its Azure cloud platform. This collaboration between the two companies has been ongoing for some time and has proven instrumental in deploying advanced AI systems. It is expected that the Orion model will be available to Microsoft’s cloud customers by as early as next month, further strengthening the partnership between OpenAI and Microsoft.

The Verge also shared an interesting detail from an OpenAI executive, who revealed that the Orion model is set to be significantly more powerful than its predecessors. According to the report, Orion is expected to be up to 100 times more capable than the GPT-4 model. Unlike GPT-4, which is known for its general-purpose capabilities in natural language processing, the Orion model is reported to be distinct from the O1 model, which is focused specifically on reasoning and problem-solving tasks. Additionally, it is believed that OpenAI has been using synthetic data derived from the O1 model to train Orion, helping to enhance its performance and precision.

One of the key takeaways from the report is that OpenAI recently completed the training process for the Orion model. This milestone coincided with a post from OpenAI CEO Sam Altman on X (formerly known as Twitter). In the post, Altman hinted at the model’s name, saying, “Excited for the winter constellations to rise soon; they are so great.” This cryptic message was widely interpreted as a subtle reference to the model’s name, “Orion,” which is also the name of a well-known star constellation.

The upcoming launch of the Orion model is highly anticipated, with many in the AI community eager to see how it will outperform previous models and what new possibilities it will unlock for businesses and developers. As OpenAI continues to push the boundaries of AI technology, the release of Orion could mark a significant leap forward in the capabilities of machine learning models.