GitHub accidentally leaked OpenAI’s GPT-5 launch

GitHub inadvertently revealed the launch of OpenAI’s upcoming GPT-5 models in a now-deleted blog post. The post, which was quickly taken down, claimed that GPT-5 is now generally available through GitHub’s model offerings. An archived version of the post confirmed that GPT-5 will be released in four distinct variants, each designed to serve different use cases. According to the leaked content, GPT-5 brings “major improvements in reasoning, code quality, and user experience.” It also features “enhanced agentic capabilities,” enabling it to handle complex coding tasks with minimal user input. The model is expected to significantly boost productivity for developers by autonomously planning, executing, and refining code. This revelation aligns with earlier reports suggesting that OpenAI was preparing to roll out GPT-5 in early August, including smaller versions like mini and nano models accessible via API. Multiple OpenAI employees had recently hinted at a major announcement, and CEO Sam Altman shared cryptic messages over the weekend, fueling speculation. The leak comes just ahead of OpenAI’s official event, scheduled for today at 10 AM PT / 1 PM ET, which the company has promoted as a “LIVE5TREAM.” The event is expected to mark the formal debut of GPT-5, along with updates on its broader AI roadmap. Adding to the momentum, OpenAI also released two new GPT-OSS open-weight models earlier this week—one of which is small enough to run locally on a standard personal computer. These open models are part of OpenAI’s growing effort to expand access to its AI technology while maintaining control over its most advanced systems. With GPT-5 poised to deliver substantial upgrades in performance and autonomy, the launch represents a pivotal moment in the evolution of generative AI, particularly for developers and technical users.