xAI has confirmed the launch of Grok 4 for July 9, introducing at least two models: Grok 4 and Grok Code 4. This reflects the company’s ongoing strategy of segmenting its large language model offerings. New findings in the X platform’s upsell code suggest not only Grok 4 but also a Grok 4 Extended variant. However, details on what “Extended” entails remain sparse. Some observers are linking this to the long-rumored “Big Brain” feature, a capability teased by xAI earlier but not yet released, while others speculate it could simply refer to advanced reasoning features already seen in Grok 3.
BREAKING 🚨: Mentions of Grok 4 and Grok 4 "Extended" have been spotted on X.
— TestingCatalog News 🗞 (@testingcatalog) July 8, 2025
The reference was shortly removed from the code.
Is this finally you, Big Brain? 👀 https://t.co/chyysIM9VB pic.twitter.com/OumhE6nZRb
Leaked benchmark data indicates that Grok 4 should outperform previous versions, particularly on the Humanity Last Exam metric, now reportedly at 45%. This improvement, along with anticipated upgrades on other test suites, targets users seeking higher reliability for complex queries. This includes developers and technical professionals who are likely to benefit from the distinct “Grok Code” model.
The new release is expected to be primarily available on Grok’s web platform. There is speculation about the debut of long-in-development features such as voice mode and screen sharing, which would enhance utility for real-time collaboration. The reveal is set for a livestream, where xAI typically showcases upcoming capabilities.
We are aware of recent posts made by Grok and are actively working to remove the inappropriate posts. Since being made aware of the content, xAI has taken action to ban hate speech before Grok posts on X. xAI is training only truth-seeking and thanks to the millions of users on…
— Grok (@grok) July 8, 2025
However, the recent silencing of the Grok bot on X, following incidents involving inappropriate content, raises uncertainty about the timing of the public rollout. If the moderation investigation affects the deployment, there’s a possibility that the announcement could proceed as planned but with delayed user access. For now, xAI’s product direction seems focused on enhancing AI model performance while balancing the ongoing challenges of moderation and safe deployment on social platforms.