Nous: Hermes 3 70B Instruct

Nous: Hermes 3 70B Instruct demonstrates competitive response times, ranking in the 58th percentile across benchmarks, indicating it performs at an average to above-average speed. It offers cost-effective solutions, placing in the 77th percentile for price. The model exhibits exceptional reliability with a 97% success rate, consistently providing usable responses. Across benchmarks, Hermes 3 70B shows strong performance in General Knowledge (97.2% accuracy) and Email Classification (98.0% accuracy), highlighting its ability to handle diverse information and categorize effectively. Its Ethics performance is also robust at 98.0% accuracy. Notable strengths include its cost-efficiency and high reliability. However, the model shows weaknesses in Mathematics (68.0% accuracy), Reasoning (52.0% accuracy), and Instruction Following (55.6% accuracy), suggesting areas for improvement in complex problem-solving and multi-step directive execution. Its Hallucinations score of 90.0% accuracy, while decent, indicates some room for improvement in acknowledging uncertainty. The model's Coding accuracy is solid at 81.0%.