Musk Releases AI Model Grok 3 And Claims It Outperforms Rivals

Elon Musk and team members of his artificial intelligence company, xAI, announced the release of Grok 3 this Tuesday, the newest and most advanced model of their chatbot. They claimed the new AI model can outperform frontier models like ChatGPT and DeepSeek.

xAI’s team and Musk announced the new model during a livestream shared on the social media platform X.

https://t.co/HTK4u2aYRu

— Elon Musk (@elonmusk) February 18, 2025

“We’re very excited to present Grok 3, which is, we think, an order of magnitude more capable than Grok 2 in a very short period of time,” said Musk.

The team explained a few of the challenges they faced—they had to delay the release due to multiple situations—and built a large data center in Memphis to train Grok 3, with 200,000 GPUs. According to TechCrunch, Grok 3 had been promised for 2024 but missed the deadline.

In January, xAI launched an iOS app for its chatbot Grok in the United States, India, and Australia.

Musk and his team showed how the new AI model outperformed GPT-4o, DeepSeek V-3, Claude 3.5 Sonnet, and Gemini-Pro in benchmarks such as AIME and GPQA.

They also announced a research feature “DeepSearch”—similar to OpenAI’s deep research feature to perform complex tasks and research online—considering data from the social media platform X as well.

Starting today, Premium X subscribers will get access to Grok 3, and xAI announced an upcoming separate subscription only to access the reasoning model.

A few experts have already shared thoughts and reviews of the new reasoning model. Andrej Karpathy, Perplexity’s CEO, said on X he tested the tool with multiple queries to analyze its potential.

“I was given early access to Grok 3 earlier today, making me I think one of the first few who could run a quick vibe check,” wrote Karpathy. “The impression overall I got here is that this is somewhere around o1-pro capability, and ahead of DeepSeek-R1, though of course we need actual, real evaluations to look at.”