xAI: Grok 4

xAI's Grok 4, released July 9, 2025, is a powerful reasoning model featuring a 256k context window, parallel tool calling, and multimodal input capabilities. While it demonstrates exceptional reliability with a 100% success rate across all benchmarks, its performance is characterized by longer response times, ranking in the 13th percentile for speed, and premium pricing, placing it in the 3rd percentile for cost. Grok 4 excels in several critical areas, achieving perfect accuracy in Hallucinations, General Knowledge, Email Classification, and Reasoning benchmarks. Notably, it is the most accurate model at its price point and speed for these categories. It also shows strong performance in Mathematics (94.5% accuracy, 93rd percentile) and Coding (93.0% accuracy, 83rd percentile). Its primary weakness appears in Instruction Following, where it achieved 78.0% accuracy, placing it in the 91st percentile but still indicating room for improvement compared to its other high-performing categories. The model's high accuracy across core reasoning and knowledge tasks, despite its higher cost and slower speed, positions it as a robust choice for applications prioritizing precision and reliability.