Elon Musk's claim that Grok excels at coding is backed by IOI Benchmark data showing Grok 3 achieving 26.2% accuracy at a $3.00/$15.00 input/output cost, outperforming GPT-5's 26.0% at $1.25/$10.00, though GPT-5's lower latency (2608s vs. 4265s) suggests trade-offs in real-world speed.The IOI Benchmark, rooted in the International Olympiad in Informatics, tests AI on complex algorithmic tasks, revealing Grok 3's edge in coding precision, a finding echoed by a 2025 study from Epoch AI that highlights Grok's strength in software optimization tasks over rivals like GPT-5.Musk's assertion comes amid xAI's legal challenge against Apple for alleged App Store bias favoring ChatGPT, a rivalry intensified by a recen t BBC report (August 8, 2025) where OpenAI's o3 model outperformed Grok 4 in an AI chess tournament, hinting at broader competitive tensions in AI development.
