Anthropic: Claude Opus 4.6

Anthropic's Claude Opus 4.6 demonstrates moderate speed performance, ranking in the 38th percentile across various benchmarks. Its pricing tends to be at premium levels, positioned in the 8th percentile. A standout feature is its exceptional reliability, achieving a 100% success rate across all benchmarks, indicating minimal technical failures and consistent evaluable responses. The model exhibits strong performance in specialized areas, particularly coding (95.0% accuracy, 91st percentile) and mathematics (95.0% accuracy, 86th percentile), aligning with its description as a powerful tool for professional tasks. It also excels in instruction following (75.0% accuracy, 81st percentile) and reasoning (94.0% accuracy, 78th percentile), showcasing its capability for complex, multi-step operations. Opus 4.6 achieves perfect accuracy in Ethics, highlighting its robust moral reasoning. While its general knowledge (99.5% accuracy, 69th percentile) and hallucination resistance (98.0% accuracy, 65th percentile) are solid, its email classification accuracy (97.0%, 39th percentile) is comparatively lower, suggesting a potential area for refinement. Overall, Opus 4.6 is a highly reliable and capable model, particularly for demanding technical and knowledge-intensive workflows, despite its premium cost and moderate speed.