xAI: Grok 4.20 Beta

Grok 4.20 Beta, xAI's latest flagship model, demonstrates strong performance across several key metrics. It consistently ranks among the fastest models, placing in the 92nd percentile for speed across eight benchmarks. Its pricing is moderate, positioned in the 22nd percentile, offering a balanced cost-to-performance ratio. The model exhibits exceptional reliability with a 99% success rate, indicating minimal technical failures and consistent response delivery. In terms of benchmark performance, Grok 4.20 Beta excels in Ethics, achieving perfect 100% accuracy, making it the most accurate model at its price point and among models of similar speed. It also shows impressive accuracy in General Knowledge (99%) and Mathematics (94%), often being the most accurate among models of comparable speed. Its Coding capabilities are strong at 92% accuracy. However, its performance in Hallucinations (90% accuracy, 38th percentile) and Instruction Following (57% accuracy, 53rd percentile) suggests areas for improvement, particularly given the description of "lowest hallucination rate on the market" and "strict prompt adherence." While its Reasoning accuracy is moderate at 74%, the model's agentic tool calling capabilities, as highlighted in its description, are not directly reflected in these benchmarks.