Grok-4 vs. ChatGPT-5: Musk Claims Victory with New Benchmarks

Elon Musk has once again stirred the AI world, making a bold claim against OpenAI and Microsoft shortly after the ChatGPT-5 release. He asserts that his Grok-4 Heavy model from xAI already outperforms its new competitor. The Benchmark Battle According to Musk, the numbers speak for themselves: Grok-4 reportedly scored 15.9% on the Arc-AGI2 test, while ChatGPT-5 achieved 9.9%. He also noted that his model was already “smarter” two weeks before the GPT-5 launch, a sentiment he claims is echoed in positive user feedback. ...

13 August, 2025 · 2 min · 259 words · Yury Akinin