Dark Mode Light Mode

Keep Up to Date with the Most Interesting News

By pressing the Subscribe button, you confirm that you have read and are agreeing to our Privacy Policy and Terms of Use
Follow Us

Keep Up to Date with the Most Interesting News

By pressing the Subscribe button, you confirm that you have read and are agreeing to our Privacy Policy and Terms of Use
Strawberry Bomb Cake: A Sweet Explosion of Flavor
Musk’s Grok-3 Unveiled! 10x Improvement Surpasses DeepSeek, The Global AI Battle Begins
Apple’s iPhone 16e: Brilliant Budget Phone or Feature-Stripped Disaster?

Musk’s Grok-3 Unveiled! 10x Improvement Surpasses DeepSeek, The Global AI Battle Begins

On February 18th, Elon Musk made his usual late entrance. After a 20-minute delay, he launched the highly anticipated flagship AI model from xAI: Grok-3. The theme of the livestream was “Our Mission is to Understand the Entire Universe.” Musk claimed that Grok-3 is “Earth’s smartest AI.” He also mentioned that its performance is 10 times better than its predecessor, Grok-2.

What Makes Grok-3 Stand Out?

Grok-3 was trained in a massive super data center located in Memphis. It used 200,000 GPUs and took just 122 days to build. The engineering team also revealed that Grok-3 is part of a model family. It includes a lightweight version called Grok-3 Mini. The Mini sacrifices some accuracy for improved speed.

In several tests, Grok-3 shattered records and showed impressive performance:

Advertisement

  • Mathematics (AIME ’24): Grok-3 scored 96 points, outperforming DeepSeek-R1 (87 points) and GPT-4.
  • Science (GPQA): Grok-3 led in graduate-level knowledge questions.
  • Programming (LCB): It generated complex code faster than Claude 3.5 and Gemini 2.0268.
  • Grok-3 also topped the Chatbot Arena leaderboard with 1402 points. It became the first model to surpass 1400 points, beating OpenAI, Google, and Anthropic.

Unprecedented Achievements in AI

At the AIME 2025 test, Grok-3 Reasoning Beta and Grok-3 Mini Reasoning scored over 90 points. They beat other competitors in reasoning challenges.

Grok-3 demonstrated its versatility by generating a full game code for Tetris × Bejeweled, complete with 3D effects. It also tackled the Riemann Hypothesis, a famous unsolved problem. The AI displayed its ability to process complex thinking steps.

Introducing the “DeepSearch” Mode

Grok-3 features a DeepSearch mode. It integrates research, brainstorming, and code debugging into one solution. Another innovative feature is the “thought chain” mechanism. It supports Test-Time Compute, which allows for deeper problem-solving. In Big Brain mode, the AI’s problem-solving depth increased by 263%.

How to Access Grok-3

However, Earth’s smartest AI isn’t available to just anyone.

X Premium+ users can unlock full access to Grok-3 immediately. Those with a SuperGrok subscription can enjoy unlimited image generation and deep search quotas. The subscription costs $30 per month.

Musk promised that once Grok-3 becomes stable, the Grok-2 model will be open-sourced. He also said, “The speed of innovation at xAI will completely rewrite the AI competition landscape!”

The Industry Reaction

The industry is already reacting. DeepSeek is open-sourced. ChatGPT is entering new territories. Will Grok-3 make a big splash? Only time will tell.

Keep Up to Date with the Most Interesting News

By pressing the Subscribe button, you confirm that you have read and are agreeing to our Privacy Policy and Terms of Use
Previous Post

Strawberry Bomb Cake: A Sweet Explosion of Flavor

Next Post

Apple’s iPhone 16e: Brilliant Budget Phone or Feature-Stripped Disaster?

Advertisement