± mixed
Qwen3.8-27B has "excellent capabilities" in coding, professional work, research and long-horizon agentic tasks, and matches the performance of a model ten times its size.
The size claim holds on the independent index: at 52, Qwen3.8-27B ties DeepSeek's V4-Flash, a 284-billion-parameter model roughly ten times larger. The capability claim is not measurable as worded, and the same index shows the model spending 160 million tokens to complete its battery against a 43 million median for its size class.