Elon Musk has just shed some light on what’s coming: xAI is skipping Grok 3.5 in favour of a direct launch of Grok 4 after July 4. This extra training on their specialised coding model promises a product that’s been given a proper tune-up.
Grok 4 is set to step up its game in coding assistance at a time when the competition is fierce. Recent looks at the codebase show that xAI is adding a native code editor to the Grok web interface – a clear nod to the functionality of VSCode – so that the AI can actually write, tweak, or debug code on its own. This move towards what some call “agentic coding” means that instead of just offering suggestions, Grok will be ready to actively work on your code within an integrated development environment.
Whether you’ll access this new tool via the web or through a standalone desktop client is still up in the air, but the aim is clear: to keep Grok competitive against other AI developer assistants that offer deep integration.
Built as a set of specialised models—including those for reasoning and search—Grok is gearing up to add a robust coding component to its repertoire. Although the launch is hinted to be shortly after July 4, it wouldn’t be surprising if delays crop up, especially with GPT-5 and Google’s upcoming Gemini Deep Think upgrade on the horizon.
If you’ve ever struggled with clunky developer tools, Grok 4 might just offer a more seamless and integrated experience. It’s part of xAI’s broader plan to not just keep pace with but also outdo its rivals in the fast-moving world of generative AI.
