Virtuoso Large
Arcee AI · 2025-05-05 · 72.7B parameters
Virtuoso Large is Arcee AI's dense 72.7B-parameter general-purpose model, fine-tuned from Qwen2.5-72B with Arcee's DistillKit and MergeKit pipeline. First released as a hosted 128K-context model for cross-domain reasoning and creative writing, it was later published as open weights under the Qwen License Agreement for domain-specialist analysis and high-fidelity content generation.
Benchmark scores
| Benchmark | Score |
|---|---|
| SnakeBench Average Apples | 0.2 |
| SnakeBench Best Apples | 1.0 |
| SnakeBench Rating | 16.5 |
| SnakeBench Total Apples | 5.0 |
| SnakeBench Win Rate | 38.1 |