Artificial Analysis has published benchmark results for the Qwen3.8-27B model, comparing its performance directly with DeepSeek V4 and GPT-5.6 Luna Max.
The analysis positions Qwen3.8-27B as being on par with these other large language models in terms of capability.