OpenAI’s GPT-5 set to launch in early August with advanced reasoning and mini versions: Report

OpenAI is gearing up for the anticipated debut of its next AI model, GPT-5, potentially arriving as soon as August, according to a report by The Verge. This release promises to include both 'mini' and 'nano' versions accessible via API, broadening the range of tools available for developers and enterprises. Reports indicate that Microsoft engineers have been diligently preparing server capacity since May. Although the initial launch timeline faced delays due to additional testing, OpenAI's CEO, Sam Altman, hinted at the imminent arrival of GPT-5 in a recent podcast, stating, "We are releasing GPT-5 soon." Altman shared a personal anecdote from his testing experience, expressing a mix of awe and humility as he described how the model flawlessly answered his queries. The core GPT-5 model is expected to incorporate advanced reasoning capabilities similar to those found in OpenAI’s prior o-series models, particularly the o3. This integration aims to enhance user experience in ChatGPT by streamlining the selection process between different model types. The unified system will also be accessible via API, ensuring wider usability. While the mini version will be available on both ChatGPT and API, the nano version will be exclusively for API access. The launch of GPT-5 signifies a significant move toward what OpenAI envisions as artificial general intelligence (AGI). Altman has previously noted that the model may not immediately achieve peak capabilities, but the consolidation of model families is part of OpenAI’s strategy to optimize AI performance and applications. The pursuit of AGI remains a critical juncture for OpenAI, especially under the current terms of collaboration with Microsoft, which entails relinquishing rights to future revenue and model access upon the formal declaration of AGI. Recently, both companies have been engaged in discussions to revisit their partnership agreements, as OpenAI aims to transition part of its operations into a for-profit structure. In the meantime, another model release is on the horizon. OpenAI is reportedly set to unveil an open-weight language model by the end of July, which is said to resemble the o3 mini and will feature advanced reasoning capabilities. This forthcoming model will mark the first open-weight release by OpenAI since GPT-2 in 2019 and will be distributed across platforms such as Azure, Hugging Face, and other cloud services. Altman confirmed a slight delay due to safety evaluations and high-risk area reviews, but sources indicate that OpenAI is eager to launch this model ahead of GPT-5, potentially as a strategic response to competitors.