Next-generation reasoning model that runs locally in-browser
Generate high-resolution images from text prompts