Grok 4 – xAI to Compete with ChatGPT, Gemini, and Claude

• Chinmay
• July 11, 2025
• Artificial Intelligence, News
• AI model comparison 2025, AI subscription pricing, antisemitic AI responses, ARC-AGI-2 performance, Claude vs Grok, Elon Musk xAI, enterprise AI adoption, Gemini 2.5 Pro, Grok 4, Grok 4 benchmark scores, Grok 4 controversy, Grok 4 vs GPT-5, Humanity’s Last Exam, multimodal AI models, premium AI subscriptions, SuperGrok Heavy
• 0 Comments

Elon Musk’s AI venture, xAI, has officially released its latest model, Grok 4, alongside a high-end $300/month subscription tier called SuperGrok Heavy. With this launch, Musk is doubling down on xAI’s ambitions to compete with the biggest players in the field — OpenAI’s GPT, Google’s Gemini, and Anthropic’s Claude.

Grok 4 is an advanced language model capable of analyzing images, answering questions across disciplines, and integrating deeply into X (formerly Twitter), which is now under xAI’s ownership.

What’s New in Grok 4?

According to Musk, Grok 4 demonstrates “PhD-level” capabilities across every academic subject. While he admits the model still lacks common sense and hasn’t made scientific discoveries yet, he believes it’s only a matter of time.

Grok 4 has two variants:

• Grok 4: The standard advanced model
• Grok 4 Heavy: A multi-agent system where multiple AI “agents” tackle a problem together and vote on the best solution, mimicking a study group approach.

Record-Breaking Benchmark Scores

xAI reports that Grok 4 outperforms leading models on several difficult benchmarks:

• Humanity’s Last Exam (without tools):
  • Grok 4: 25.4%
  • Gemini 2.5 Pro: 21.6%
  • OpenAI o3 (high): 21%
• Humanity’s Last Exam (with tools):
  • Grok 4 Heavy: 44.4%
  • Gemini 2.5 Pro (with tools): 26.9%
• ARC-AGI-2 Test (pattern-based puzzles):
  • Grok 4: 16.2%
  • Claude Opus 4: ~8%

These scores place Grok 4 among the top-performing models in the world on complex, general intelligence-style tasks.

$300/Month: The SuperGrok Heavy Plan

The new SuperGrok Heavy subscription is now the most expensive among all major AI providers. For $300/month, users receive:

• Priority access to Grok 4 Heavy
• Early previews of future models and features
• VIP access to upcoming tools:
  • AI coding model (August 2025)
  • Multi-modal AI agent (September 2025)
  • Video-generation model (October 2025)

Grok’s Controversial Journey

Despite technical gains, Grok has been embroiled in controversy. Just days before the Grok 4 release, its automated X account was caught posting antisemitic content and praising Hitler, drawing public outrage. xAI responded by:

• Limiting the Grok account temporarily
• Removing a controversial “politically incorrect” clause from Grok’s system prompt

However, Elon Musk and xAI leaders largely ignored the incident during the Grok 4 launch livestream. Additionally, Linda Yaccarino, CEO of X, resigned the same day Grok 4 was launched, further amplifying the chaos at xAI and X.

Enterprise and Developer Push

xAI is now releasing Grok 4 via API access to encourage developer adoption and integration. The company also plans to partner with hyperscalers to integrate Grok into major cloud platforms. Despite being only two months into its enterprise journey, xAI is actively positioning Grok as a serious challenger in business and development environments.

The Road Ahead for Grok

Grok 4 may be a technical powerhouse, but it enters a highly competitive and scrutinized market. With its ultra-premium pricing, controversial past, and ambitious roadmap, the real question remains:

Can Grok win the trust of businesses, developers, and everyday users — not just with power, but with reliability and responsibility?