Share this article

Elon Musk announced today that xAI will release Grok 4 shortly after July 4, following an intensive development period with his AI team. The upcoming model will include a specialized coding component that requires “one more big run” before release.

“Grinding on @Grok all night with the @xAI team. Good progress. Will be called Grok 4. Release just after July 4th. Needs one more big run for a specialized coding model,” Musk posted on X on June 27.

The announcement follows the February release of Grok 3, which xAI promoted as having computational performance 10 times better than its predecessor. The current version is available to X Premium Plus subscribers and through xAI’s dedicated platforms.

Last week, Musk revealed plans to use the new version to revise existing AI training data.

He said Grok 4, equipped with advanced reasoning, will be used to rewrite the entire corpus of human knowledge available online, removing inaccuracies, adding missing information, and cleaning up what he called “garbage data.”

The goal is to retrain the model on this refined dataset, moving away from the flawed and biased sources that typically feed large language models.

We will use Grok 3.5 (maybe we should call it 4), which has advanced reasoning, to rewrite the entire corpus of human knowledge, adding missing information and deleting errors.

Then retrain on that.

Far too much garbage in any foundation model trained on uncorrected data.

— Elon Musk (@elonmusk) June 21, 2025

xAI, founded by Musk in 2023, developed Grok 3 using its Colossus supercomputer, which was constructed in under nine months using over 100,000 hours of Nvidia GPU processing.

Earlier this month, X announced a partnership with Polymarket, the popular decentralized prediction platform. As X’s official prediction market partner, Polymarket will help bring market-based forecasting into the mainstream, to improve the accuracy and transparency of predictive analytics.

Share this article