Skip to content

model

Qwen3.5 122B A10B NVFP4

NVFP4-quantized build of a 122-billion-parameter Qwen mixture-of-experts language model with roughly 10 billion parameters active per token.

Current clusters