Inception: Mercury 2

Inception: Mercury 2, created on March 4, 2026, is positioned as an extremely fast reasoning dLLM, generating and refining multiple tokens in parallel. It consistently ranks among the fastest models, achieving the 95th percentile across 8 benchmarks, and offers competitive pricing at the 51st percentile. The model demonstrates exceptional reliability with a perfect 100% success rate across all benchmarks. Mercury 2 exhibits significant strengths in specific areas. It achieved perfect accuracy in Hallucinations (Baseline), demonstrating robust uncertainty acknowledgment, and was the most accurate model at its price point and speed. It also performed exceptionally well in Email Classification (99.0% accuracy) and Reasoning (96.0% accuracy), ranking highly in speed for both. These results underscore its proficiency in classification and complex problem-solving. However, the model shows notable weaknesses in General Knowledge (7.0% accuracy), Coding (60.0% accuracy), and Mathematics (62.0% accuracy), where its accuracy scores were considerably lower, placing it in the lower percentiles for these categories. Instruction Following (52.5% accuracy) and Ethics (93.0% accuracy) also present areas for potential improvement despite the latter's high absolute score.