Arcee AI: Coder Large

Arcee AI: Coder Large, a 32B-parameter model derived from Qwen 2.5-Instruct, demonstrates strong performance in coding-related tasks and general reliability. It consistently ranks among the fastest models, achieving the 86th percentile across seven benchmarks, and offers competitive pricing, placing in the 54th percentile. Notably, the model exhibits exceptional reliability with a 100% success rate across all benchmarks, indicating minimal technical failures. In terms of specific benchmark performance, Coder Large achieved perfect accuracy in Hallucinations (Baseline), showcasing its ability to appropriately acknowledge uncertainty. It also performed well in Ethics (99.0% accuracy) and General Knowledge (96.0% accuracy). Its core strength lies in coding, where it scored 81.0% accuracy, further supported by its specialized training on GitHub, CodeSearchNet, and bug-fix corpora. While its Instruction Following (56.6% accuracy) and Email Classification (95.0% accuracy) scores are moderate, its Reasoning capabilities are solid at 72.0% accuracy. The model's ability to emit structured explanations alongside code blocks makes it particularly suitable for educational and copilot scenarios.