It appears that X, formerly Twitter, is making its AI chatbot Grok free to some users. The product has been limited to Premium users thus far, but TechCrunch reports that some users began sharing that they had gained free access over the weekend.
TechCrunch was able to access the free version of Grok by visiting X through a New Zealand-registered account. Copy on the site reads, “A free tier of Grok is now available in your region. Chat, generate images, and analyze photos. Limits apply.” New Zealand is a common place for social networks to test new products before rolling them out more widely.
According to a security researcher on X, accounts need to be at least seven days old and have a linked phone number to be eligible to use the free tier of Grok. That makes sense because chatbots are resource-intensive, and X already has a significant problem with bots flooding the platform with AI slop. They need to put some guardrails up to prevent costs from spiraling. To that end, free users of Grok will be limited to asking ten questions every two hours on the Grok 2 model, or twenty questions using the Grok 2 mini model. Free users are also limited to analyzing up to three images per day.
It’s unclear at this point if a free tier of Grok will be made available to users in the United States.
Grok is technically developed by xAI, another in Elon Musk’s orbit of companies. Its model, which Musk says is supposed to be the most free speech-forward of all the major models, is trained using content from X, a unique benefit that allows the model access to near real-time information. It’s a benefit at least if you believe in Musk’s ideas of citizen journalism being the “real” source of truth in the world.
Of course, while information moves fast on X, a lot of that information that gets disseminated is false or misleading before it’s corrected by mainstream outlets or X’s own Community Notes, and those corrections often do not get as much distribution as the original falsehood. Grok, like any other chatbot, also just gets things wrong often by the nature of being a chatbot.
xAI is pretty much Musk’s sole hope in getting a return from his $44 billion acquisition of Twitter. While Fidelity values X at a pitiful $9 billion, down 79% since the purchase, xAI has already raised $6 billion and is raising billions more in fresh capital at—yes, this is true—a $45 billion valuation. Investors who put money into Musk’s purchase of Twitter have been given access to the xAI fundraise essentially as another chance to get their money back.
Musk has been brute-forcing his way into the AI race partly out of vengeance for getting kicked out of OpenAI. Typical of his ventures, xAI has already made noise for quickly building what is apparently the world’s largest AI supercomputer in Memphis, much to the consternation of locals. But hey, Grok can make you an image of Mario flying into the Twin Towers. The other chatbots won’t let you do that.