PrismML: Ternary Bonsai 2 27B

PrismML: Ternary Bonsai 2 27B, a 27B-parameter reasoning model derived from Qwen3.8-27B, demonstrates exceptional reliability with a 100% success rate across all benchmarks, consistently providing usable responses. It performs among the fastest models and offers highly competitive pricing, ranking in the Infinityth percentile for both speed and cost across five benchmarks. While excelling in General Knowledge (98.5% accuracy), Email Classification (99.0% accuracy), Reasoning (96.0% accuracy), and Ethics (99.0% accuracy), the model exhibits a critical weakness in Instruction Following, achieving 0.0% accuracy. This suggests a significant limitation in processing and executing complex, multi-layered instructions despite its strong performance in other cognitive areas. Its high accuracy in classification, reasoning, and ethical scenarios, coupled with its impressive speed and cost-efficiency, positions it as a strong contender for tasks not heavily reliant on intricate instruction adherence. The model's long context window of 262K tokens further enhances its utility for extensive data processing.