Qwen3.5 122B A10B

Open-weight Qwen3.5 native vision-language MoE model (122B total, 10B active) with hybrid linear attention.

Select Provider