Elon Musk xai Startup has unveiled the latest version of its flagship model of artificial intelligence (AI), Grok 4.
In a livestream On X, Musk boasted of the model Again Cross simultaneously on the impact on humanity if round wrong.
“This is the most intelligent AI in the world,” said Musk when he was surrounded by members of his team XAI. “In some ways, it’s terrifying.”
He compared Grok 4 to a “super-genius child“In which the” good values ”of veracity and a sense of honor must be instilled in order for society to benefit from its advances.
Musk admitted to being “worried”, saying that “it is somewhat annoying to have the intelligence that is much greater than ours create, and will it be bad or good for humanity?”
The owner of Xai concluded that “most likely, it will be good”.
Musk said that Grok 4 is designed to perform at the “post-diviple level” in a lot subjects simultaneously, which no one can do. He can also manage images, generate realistic visuals and fight against complex analytical tasks.
Musk claims that Grok 4 would perfectly mark the SAT and Graduate exams at the level as Gree even Without seeing the questions in advance.
In addition to the model version, XAI has introduced Supergrok Heavy, a price subscription level of $ 300 per month. A standard grok 4 level is available for $ 30 monthly, And The basic level is free.
OPENAI,, Google,, Anthropic And Perplexity also unveiled higher prices: Chatgpt Pro, at $ 200 per month; Gemini Ultra, $ 249.99 per month; Claude Max, at $ 200 per month; And Perplexity Max, for $ 200 per month.
See also:: Elon Musk Startup XAI launches the application offering access to Grok Chatbot
Turbulent week for grok and x
The launch of Grok 4 follows a turbulent week marked by anti -Semitic content generated by Grok 3 and the Resignation of Linda YaccarinoThe CEO of X.
GROK 4 is published in two configurations: the standard GROK 4 and the premium “heavy” version.
The heavy model includes a multi-agent architecture capable of collaborative reasoning on difficult problems.
The model Demonstrates advances in multimodal treatment, faster reasoning and an improved User interface. According to XAI, Grok 4 can solve complex mathematical problems, interpret images – including scientific visuals such as collisions of black holes – and carry out predictive analyzes, such as estimation of the chances of a team to win a championship.
The reference data shared by XAI show that Grok 4 has outperformed the previous models on tests such as the last examination of humanity.
Xai described an aggressive roadmap for the rest of 2025:: launch of a Coding specific In August, a multimodal agent in September and a model capable of generating video in October.
The release of Grok 4 intensifies competition among the main companies of AI. Openai should deploy GPT– 5 Later this summer, while Google continues grow His Gemini series.
Find out more:
