@elonmusk: Grok foundation model V9-Medium (1.5T) has finished training. Evals look good. A lot of Cursor data was added in supple…

X AI KOLs Following Models
grok foundation-model training ai xai large-language-model

Summary

Elon Musk announced that the Grok foundation model V9-Medium (1.5T parameters) has finished training with strong evaluations, and will be publicly released in 2-3 weeks after fine-tuning and reinforcement learning.

Grok foundation model V9-Medium (1.5T) has finished training. Evals look good. A lot of Cursor data was added in supplementary training and there is more to come. Fine-tuning is underway and reinforcement learning begins in a few days. 2 to 3 weeks to public release. This will be a major improvement over the 0.5T v8-small that currently serves all Grok production traffic, especially for difficult coding tasks.
Original Article
View Cached Full Text

Cached at: 05/25/26, 06:32 AM

Grok foundation model V9-Medium (1.5T) has finished training. Evals look good. A lot of Cursor data was added in supplementary training and there is more to come.

Fine-tuning is underway and reinforcement learning begins in a few days. 2 to 3 weeks to public release.

This will be a major improvement over the 0.5T v8-small that currently serves all Grok production traffic, especially for difficult coding tasks.

Similar Articles