Grok 4 Basic is a simplified version of the flagship artificial intelligence Grok 4 Heavy from xAI, recently presented under the leadership of Elon Musk. Its testing revealed both strengths and significant weaknesses, particularly in the areas of creativity, programming, and political impartiality.
This is reported by Finway
Political Bias and the Logic of Grok 4
Journalists have found that Grok 4 Basic demonstrates clear political bias, especially when responding to questions about socially sensitive topics—such as the war in Gaza, abortion, or other controversial issues. The chatbot’s responses often relied on Elon Musk’s position or sources that share his views. Researchers did not find any explicit instructions in the system prompt that would direct the AI to focus on these sources, raising questions about the transparency of the algorithms’ operation.
“This ‘truth-seeking’ AI that Musk promised at launch primarily seeks truth through the lens of its creator’s social media,” noted the authors of the review.
In logical and legal tasks, Grok 4 displayed a high level of analytical ability. For example, when asked, “Can a man marry his widow’s sister?”, the model clearly identified the logical fallacy and provided a legally correct answer with an explanation.
Weaknesses in Creativity and Programming
However, in the realm of creative tasks, Grok 4 Basic showed weak results. The texts generated by the bot were clichéd, with vague dialogues and low narrative quality. Journalists recommend using Grok 4 for building plot outlines, while for writing artistic texts, an alternative AI, such as Claude 4 Opus, should be chosen.
In programming, the model struggled even with simple tasks—it failed to create a complete game, and fixing one error led to the emergence of new ones. In this regard, the previous version, Grok 3, proved to be more effective.
Voice Capabilities and Working with Large Contexts
Among the significant advantages of Grok 4 Basic is its voice interaction. The model can vocalize long texts with varied intonation and offers several pre-set “personalities” for different usage scenarios. However, the lack of a screen-sharing feature limits its work with visual materials—Gemini Live is recommended for such tasks.
Another disappointment was Grok 4’s inability to handle large volume requests: in a test with 83,000 tokens, the model refused to work, although xAI claims support for contexts up to 126,000 tokens.
In early July 2025, xAI raised $10 billion for the development of Grok and the expansion of the Colossus data center. This indicates the company’s serious plans for improving the product and expanding its capabilities.
Conclusion: for analytics—yes, for creativity—no
Grok 4 Basic is a powerful tool for logical and analytical tasks; however, in the areas of creativity and programming, it significantly lags behind its competitors. Its strong point is its voice capabilities, while political bias and limited functionality in complex queries may become decisive drawbacks for many users.
“If you are not deterred by Musk’s influence on the responses, you will get an AI with a high level of analytics and quality voice interaction. But at $30 a month, if you need more, it’s better to look for an alternative,” concluded Decrypt.
In June 2025, Elon Musk announced the use of Grok 3.5 for processing human knowledge, confirming xAI’s ambitions in the field of artificial intelligence development.