The Grok foundation model V9, boasting 1.5 trillion parameters, has completed its training phase, marking a significant upgrade from the previous 0.5 trillion parameter V8 model. The development team is now incorporating Cursor data through supplemental training, followed by Supervised Fine-Tuning (SFT) and Reinforcement Learning (RL). The enhanced model is expected to be released in approximately three to four weeks, promising substantial improvements in performance and capabilities.