Social network X is now offering its Grok chatbot to Premium subscribers, following Elon Musk’s recent announcement about expanding its availability to more paid members. According to the company’s support page, only those with Premium and Premium+ subscriptions in specific regions will be able to interact with the chatbot.
Last year, when xAI introduced the Grok chatbot, it was initially accessible only to Premium+ subscribers, who pay either $16 monthly or a yearly fee of $168. With this new update, users with an $8 monthly subscription can also use the chatbot.
Users have the option to engage with Grok in either “Regular mode” or “Fun mode”. Like other Large Language Model (LLM) products, Grok provides disclaimers indicating that it might not always provide accurate answers.
You might also like: Elon Musk Promises that Grok will going to be an Open Source
Some instances of this have already been observed. Recently, X introduced a new feature in Grok that presents a summary of trending news stories. Interestingly, both Jeff Bezos and NVIDIA-backed Perplexity AI also offer similar news summary features.
However, Grok takes it a step further by not just summarizing but also creating headlines. As reported by Mashable, the chatbot generated a misleading headline stating, “Iran Strikes Tel Aviv with Heavy Missiles.”
Musk appears to be aiming for more widespread adoption of the Grok chatbot to compete with other products like OpenAI’s ChatGPT, Google’s Gemini, and Anthropic’s Claude. In recent months, he has been openly critical of OpenAI’s practices. Musk even filed a lawsuit against the company in March, accusing it of deviating from its non-profit objectives. In response, OpenAI countered by seeking the dismissal of Musk’s claims and making public email exchanges between Musk and the company.
Read more: Elon Musk’s AI Startup Releases Sassy Chatbot Grok
Grok was made publicly available last month by xAI, but no information regarding the training set was made public. As noted by my colleague Devin Coldewey, there are lingering questions about whether this version is the most recent iteration of the model and whether the company will be more transparent regarding its development process and training data.