Qwen 3.8 27B scores 52 on the Artificial Analysis Intelligence Index

The Qwen 3.8 27B artificial intelligence model achieves a score of 52 on the Artificial Analysis Intelligence Index. This performance matches the score of GPT-5.6 Luna and places it just one point behind GLM-5.2 and DeepSeek V4 Pro 0813.

The achievement is notable because of the model's size relative to its competitors. The GLM model contains 753 billion parameters, and the DeepSeek model contains 1.6 trillion parameters, while the size of the Luna model is unknown but presumed to be much larger than 27 billion parameters. Despite these vast differences in scale, the Qwen model performs at a comparable level.

The model is considered exceptional, though it has a tendency to default to excessive overthinking in its responses.