Muse Glimmer 30B handles both text and images, making it a multimodal open-weight model in the 30B parameter range. It operates with a large 131K token context window, allowing it to work with lengthy inputs without truncation. Details about its specific behavioral tendencies and capability profile are limited beyond these structural characteristics.