The Qwen3.5 series 397B-A17B native vision-language model is built on a hybrid architecture that integrates a linear attention mechanism with a sparse mixture-of-experts model, achieving higher inference efficiency. It delivers...
模态
文本图像视频
输入 / 输出价格
₩965 / ₩6,143
1M
上下文
26.21万
发布日期
2026年2月16日
暂无活动数据。