OpenAI: gpt-oss-120b

OpenAI's gpt-oss-120b, an open-weight 117B-parameter MoE model, demonstrates a balanced performance profile suitable for high-reasoning and agentic production use cases. It exhibits competitive response times, ranking in the 57th percentile for speed, and offers competitive pricing, placing in the 59th percentile. A standout feature is its exceptional reliability, boasting a 97% success rate across benchmarks, indicating consistent operational stability. The model shows significant strengths in several key areas. It achieves high accuracy in Coding (93.0%, 89th percentile), General Knowledge (99.5%, 82nd percentile), Email Classification (99.0%, 88th percentile), Reasoning (96.0%, 89th percentile), and Mathematics (94.0%, 90th percentile). Notably, it performs among the fastest models in Keyword Topic Relevance Classification. Its hallucination rate is commendably low at 98.0% accuracy. However, the model shows a notable weakness in Instruction Following, with only 46.0% accuracy (44th percentile), and its Ethics performance is in the lower quartile (93.8% accuracy, 26th percentile), suggesting areas for improvement in complex ethical reasoning.