Muse Glimmer NVFP4 is a multimodal model from RadixArk's Muse family that accepts both text and image inputs, producing text outputs. It ships in NVFP4 quantization format via safetensors, suggesting it's tuned for efficient inference on compatible hardware. Details about its specific reasoning style or capability profile are limited beyond its open-weight, vision-language nature.