Elon Musk’s xAI to Release Grok 4 and Grok 4 Code AI Models After July 4 With Advanced Benchmarks and Knowledge Rewrite

Elon Musk and his xAI team are preparing to release Grok 4, the latest iteration of their AI language model, shortly after July 4, 2025. Grok 4 is designed to be the most advanced version yet, with capabilities to rewrite the entire corpus of human knowledge by correcting errors, filling gaps, and enhancing reasoning. The AI will include at least two models: Grok 4, an all-rounder in language, math, and reasoning, and Grok 4 Code, a specialized model focused on software engineering tasks. Recent leaked benchmark results suggest Grok 4 significantly outperforms previous models, achieving scores such as 35% on the Humanities Last Exam (HLE) with a 45% score when reasoning is applied, 87-88% on the General Purpose Question Answering (GPQA) benchmark, and 72-75% on the Software Engineering (SWE) benchmark for Grok 4 Code.

These benchmarks indicate a potential state-of-the-art performance level in AI language models. Musk has been personally involved in the development, working intensively with the xAI team to finalize the model. Although Grok 4 was expected to launch immediately after July 4, 2025, the release has been delayed by a few days, with Musk confirming a major upgrade to Grok is imminent. The team is focusing on improving accuracy and reliability, addressing previous issues related to biased or flawed data in AI training. The rollout of Grok 4 is highly anticipated as it aims to redefine AI capabilities in knowledge processing and coding assistance.