xAI Unveils Grok-2 and Grok-2 Mini: New AI Models with Image Generation Capabilities on X
Ibrahim Awotunde
Elon Musk’s AI company, xAI, has just launched its latest AI models, Grok-2 and Grok-2 Mini, both currently in beta.
These new models represent a significant leap forward in AI capabilities, offering enhanced reasoning and the ability to generate images directly on the X social network.
However, access to these models is currently restricted to Premium and Premium+ users on X.
xAI announced that Grok-2 and Grok-2 Mini are now available for testing on X. The company expressed enthusiasm about the progress, stating in a blog post, “We are excited to release an early preview of Grok-2, a significant step forward from our previous model Grok-1.5, featuring frontier capabilities in chat, coding, and reasoning. At the same time, we are introducing Grok-2 Mini, a small but capable sibling of Grok-2. An early version of Grok-2 has been tested on the LMSYS leaderboard under the name ‘sus-column-r.’”
In addition to these advancements, xAI is gearing up to make both models available to developers through its enterprise API later this month. This move is expected to expand the utility of Grok-2 and Grok-2 Mini, allowing businesses and developers to integrate advanced AI functionalities into their own applications.
One of the most intriguing features of Grok-2 is its ability to generate images, a capability that has already caught the attention of users on X. Early adopters have started experimenting with this feature, generating images that include political figures—a functionality that currently lacks any apparent restrictions.
With the U.S. presidential election on the horizon, it remains to be seen whether xAI will implement any guardrails to prevent the misuse of this feature for spreading misinformation.
The image generation feature of Grok-2 appears to utilize FLUX.1 by Black Forest Labs, as indicated by text prompts below sample images. Despite the excitement surrounding these new capabilities, details about Grok-2’s full range of functionalities remain scarce.
Some app researchers, like Nima Owji, suggest that Grok-2 excels in code generation, writing, and news summarization, although the first version of Grok had its fair share of issues with accurate news summaries.
The lack of limitations on Grok’s image generation could potentially make it a tool for creating misleading content, especially in a politically charged environment.
It is also unclear whether images generated by Grok are embedded with any metadata to indicate their AI origin, a feature that could be crucial in curbing the spread of misinformation.
xAI has been tight-lipped about how it plans to address these concerns. Despite reaching out for comments, the company has not provided any responses, which aligns with its trend of limited media engagement since Musk took over X.
Looking ahead, xAI has ambitious plans for Grok-2 and Grok-2 Mini. The company aims to integrate these models into AI-driven features on X, such as enhanced search capabilities, post analytics, and possibly AI-powered replies.
Additionally, xAI mentioned that it will soon release a preview of multimodal understanding, which will become a core part of Grok’s functionality on X and its API.
As Grok-2 and Grok-2 Mini continue to evolve, their impact on the X platform and beyond will be closely watched, especially as discussions around AI ethics and responsible usage intensify.
