OpenAI is already training o4 and expects “another big jump in capabilities”

OpenAI has already begun training its next reasoning model, according to statements made at the World Economic Forum in Davos. During his presentation on ChatGPT‘s roadmap to 2025, OpenAI chief product officer Kevin Weil revealed that the company is working on a successor to its yet-to-be-released o3 model, likely called o4. Weil expects this new model to bring another “another big jump in capabilities,” Weil said, adding that it took OpenAI just three months to go from o1 to o3 (technically o2), and he expects even shorter iteration cycles for future models.