xAI skips Grok 3.5 and eyes July launch for Grok 4 with new developer features

Elon Musk has recently clarified that xAI will skip the planned release of Grok 3.5, moving directly to Grok 4, with a launch anticipated sometime after July 4th. The decision appears to be driven by the need for another round of training for their dedicated coding model, which is expected to be a key focus in this upcoming version. This update follows several delays and an extended period where Grok 3.5 was expected but never delivered, likely due to xAI reassessing its position as OpenAI and Google rolled out major new models, raising the bar for large language model capabilities.

The next Grok iteration is expected to advance coding assistance, a space that is seeing increasing competition. Recent findings in the codebase suggest xAI is building a native code editor within the Grok web interface, modelled after VSCode. This will allow Grok to directly modify, write, or debug code, and indicates a move towards “agentic coding”, where the AI not only suggests code but acts more autonomously within an IDE. The exact implementation, whether this will remain web-based or expand to a dedicated desktop client, remains uncertain, but either approach would position Grok more competitively alongside other AI assistants that provide deep integration for developers.

Grok is structured as a system of specialised models like Reasoning, Search, etc., and is now expected to have a stronger coding component. While the promise of improved coding abilities is substantial, the release timing for Grok 4 is still not concrete. Musk’s post hints at a post-July 4th launch, but no specific date was committed, and further delays remain possible, especially with GPT-5 and Google’s Gemini Deep Think upgrade expected to launch soon and set new benchmarks.

For developers and tech professionals, these updates signal a push towards richer, more integrated development tools powered by AI. For xAI, this aligns with their broader strategy to stay at the frontier of generative AI, not only by matching but by trying to leapfrog features from major competitors with every new version. However, much will depend on how Grok 4 performs in practice, especially as the competitive landscape heats up in July.