Grok 3.5 Launch Seems Near as Indicators Emerge in Code
The debut of Grok 3.5, the forthcoming version of xAI’s AI model, might be imminent. Clues regarding this new iteration have been found in recent code revisions, suggesting that the official release could take place any moment now.
As reported by TestingCatalog, references to Grok 3.5 have begun to surface in the newest web version of xAI’s platform. The site also provided screenshots on X (previously known as Twitter) highlighting these mentions, although the original tweets are no longer accessible.
Alongside the version upgrade, TestingCatalog has also detected indications of a new API and a possible screen-sharing functionality for Grok’s voice mode on iOS. This would bring Grok in line with features already available from competitors such as OpenAI’s ChatGPT and Google’s Gemini Live.
In late April, xAI founder Elon Musk revealed that an initial beta version of Grok 3.5 would be offered to SuperGrok subscribers. This signaled the onset of what many anticipated to be a more extensive rollout. Recently, Musk commented on an X post regarding the forthcoming model, stating that Grok 3.5 would excel at “reasoning from first principles (truth),” indicating enhanced logical and analytical abilities.
Grok has been positioned as an “unfiltered” AI model, presenting fewer safety limitations compared to other platforms. However, this strategy has sparked controversy. A recent bug resulted in the model repeatedly mentioning “white genocide in South Africa” in reply to unrelated questions, provoking backlash and raising concerns about content moderation.
Currently, Grok 3 features options like image generation and research tools, making it competitive for general purposes. Nevertheless, it still falls behind competitors such as Google’s Gemini and Anthropic’s Claude in aspects like coding. According to Bleeping Computer, Grok 3.5 is anticipated to bridge that gap with upgraded programming skills.
With indicators suggesting an upcoming launch, Grok 3.5 could be on the verge of being accessible to the public. Once it becomes available, it will be evident how Musk’s newest AI model measures against its competitors.