Arcee AI: Caller Large

Arcee AI: Caller Large is a specialized function-calling SLM designed for orchestrating external tools and APIs, prioritizing structured JSON outputs and multi-step tool chains over next-token accuracy. The model performs among the fastest models, typically ranking in the top tier for speed (76th percentile across 5 benchmarks). It also offers competitive pricing, falling around the 50th percentile across benchmarks. Notably, Caller Large demonstrates strong reliability with a 95% success rate, indicating it consistently provides usable responses with few technical issues. In terms of benchmark performance, Caller Large exhibits a mixed profile. While it achieves a respectable 90.5% accuracy in General Knowledge, this places it in the 31st percentile, suggesting it's not a top performer in broad factual recall. Similarly, its 97.0% accuracy in Ethics is solid but ranks in the 34th percentile. A significant weakness is observed in Email Classification, where its 87.0% accuracy places it in the 11th percentile, indicating challenges in this specific classification task. Instruction Following shows moderate performance at 53.1% accuracy (55th percentile), while Coding is also a weaker area at 60.0% accuracy (25th percentile). Its core strength lies in its intended purpose: function-calling and structured output generation, which is not directly measured by these general benchmarks but is highlighted by its design and integration into Arcee Conductor. Its speed and reliability are key advantages, making it suitable for applications where consistent, fast tool orchestration is paramount.