Qwen: Qwen3.5-122B-A10B

Sobre

The Qwen3.5 122B-A10B 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. In terms of...

Benchmarks

Comparação Rápida