Elon Musk Releases Grok 4.1 AI Model to All Users: How to Use It?
Elon Musk's xAI company has released Grok 4.1, the successor to Grok 4, which was released in July. This new model boasts several improvements and new capabilities. The AI company claims that the latest version of its big language model offers enhanced emotional intelligence, improved creative writing, and less hallucinations.
The company also shared the results of internal benchmark tests for the model, highlighting its superior performance compared to Gemini 2.5 Pro and Claude 4.5 Sonnet in some metrics. It's important to note that this is a major update to the fourth-generation Grok model, not a completely new model.
Grok 4.1 Improvements and Features
In a newsroom post, xAI announced the release of the latest version of its AI model. Interestingly, Grok 4.1 was released secretly between November 1st and 14th. Users were presented with a random choice between two options, one of which was the newer model. xAI stated that Grok 4.1 was preferred by 64.78% of users over the previous model.
The new AI model is now available to everyone via Grok and x (formerly Twitter), as well as Android and iOS apps. Users don't even need to log in to their accounts if they use the model on the website. The default model is now set to automatic mode and can be explicitly selected through a separate Grok 4.1 menu.
Regarding improvements, emotional intelligence is one area where Grok 4.1 excels. Based on internal evaluations, the company stated that the model scored 1585 points on the EQ-Bench benchmark (1586 with Grok 4.1 Thinking), outperforming models like the Gemini 2.5 Pro, Claude Opus 4, and GPT-5. In real-world use cases, this means the AI model is now better able to distinguish subtle differences in prompts and better understand their tone, then generate a response that reflects the emotions.
The model's creative writing has also been improved. The company claimed that its LLM program achieved 1708.6 points in the Creative Writing v3 benchmark, surpassing Claude 4.5 Sonnet. This enhanced capability now allows Grok 4.1 to generate better responses to creative tasks, such as creating social media posts, writing a short story, etc. Finally, the company also claimed that the AI model hallucinates (generates inaccurate information) less than its predecessor.