Overall score
站内排名: #53 / 96更新时间: Jul 9, 2026置信度: 3
The Qwen3.5 Series 35B-A3B is a native vision-language model designed with a hybrid architecture that integrates linear attention mechanisms and a sparse mixture-of-experts model, achieving higher inf...
按基准分类汇总的 0-100 分表现。
Input and output token limits for this model, plus how it ranks on long-context understanding.
1 个指标
2 个指标
2 个指标
17 个指标
12 个指标
8 个指标