ᅠ‏ ᅠ ᅠ

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.

regarding usage limits: we identified and fixed a regression in our prompt caching for Grok Build which lead to: - faster token burn - slower grok build response times We have alerting and regression tests in place, we will also give out extra credits to all users.

I was in the middle of taping on my iPhone for a social video from the White House North Lawn when we heard the shots. It sounded like dozens of gunshots. We were told to sprint to the press briefing room where we are holding now.