Arcee AI: Virtuoso Large

Virtuoso-Large is Arcee's flagship general-purpose LLM, boasting 72 billion parameters. It is meticulously tuned for superior performance in cross-domain reasoning, creative writing, and enterprise-grade question answering. A key differentiator is its impressive 128k context window, inherited from Qwen 2.5, enabling it to process extensive documents like books, codebases, or financial filings in their entirety. This model is a powerhouse for applications requiring deep contextual understanding. Its advanced training regimen includes DeepSeek R1 distillation, multi-epoch supervised fine-tuning, and a final DPO/RLHF alignment stage, resulting in strong benchmarks on BIG-Bench-Hard, GSM-8K, and long-context Needle-In-Haystack tests. Enterprises leverage Virtuoso-Large as a reliable "fallback" brain in Conductor pipelines. Despite its size, aggressive KV-cache optimizations ensure low-second first-token latency on 8× H100 nodes. It supports functions and streaming, with a 131K token context window and 4K token max output. Pricing is competitive at $0.75/1.20 per 1M tokens (input/output).