Grok V9: A Quantum Leap in AI Scale
Addressing community inquiries about development timelines, Elon Musk provided a clear update on the Grok AI model's roadmap. While incremental improvements continue on the current version, the spotlight is now on the next-generation iteration.
The Path to Launch
The newly trained V9 model represents a substantial architectural advancement, scaling to 1.5 trillion parameters. Before its public debut, the model will undergo a rigorous refinement phase:
- Supplementary training using curated data from Cursor
- Comprehensive supervised fine-tuning (SFT)
- Reinforcement learning (RL) optimization
Musk indicated this finalization process requires approximately three to four weeks, setting a clear timeline for the anticipated release.
Ongoing Refinement of Current Version
Parallel to V9's development, the existing V8 model continues receiving regular enhancements. Though publicly available at 430 billion parameters, internal versions are being optimized every few days, ensuring users benefit from continuous improvements.