Qwen3.8 27B Scores 52 on Artificial Analysis' Intelligence Index — #1 of 135 Open-Weights Models, Apache 2.0, 256K Context
Artificial Analysis / Hacker News·high signal
Alibaba's newly released Qwen3.8 27B now ranks first among 135 comparable open-weights models on the Artificial Analysis Intelligence Index at 52, against a median of 9 for that cohort. It is a 27B reasoning model with a 256K context window, image input, text output, and an Apache 2.0 license permitting commercial use. The one practical caveat in the eval data: it emitted 160M output tokens across the benchmark suite versus a 43M median, so extended thinking makes it far more verbose than peers — real cost per task will not track its headline price.