xAI: Grok 3 Mini Beta

Grok 3 Mini Beta demonstrates a balanced performance profile, excelling in accuracy while maintaining moderate speed and pricing. Its speed performance places it in the 31st percentile, indicating it operates at a moderate pace compared to other models. Pricing is similarly moderate, ranking in the 33rd percentile, suggesting it offers competitive costs. A standout feature is its exceptional reliability, achieving a perfect 100th percentile, meaning it consistently provides usable responses with virtually no technical failures. The model exhibits remarkable accuracy across several benchmarks. It achieved perfect scores in Ethics, Reasoning, and General Knowledge, often being the most accurate model at its price point or speed. Its "thinking before responding" mechanism appears to contribute significantly to these high accuracy rates, particularly in reasoning-heavy tasks. In Coding, it scored an impressive 94.0% accuracy, placing it in the 95th percentile. While its Instruction Following accuracy was 71.7%, this still ranks in the 89th percentile, indicating strong capability. Email Classification showed solid performance at 98.0% accuracy. Grok 3 Mini Beta's key strengths lie in its high accuracy for complex reasoning, ethical considerations, and general knowledge, making it ideal for tasks requiring precision and deep thought. Its primary limitation is its moderate speed, which might not suit applications demanding instantaneous responses.