WeLM-Image: A Unified Image Generation Model towards Visual Intelligence
WeLM-Image combines multimodal understanding, image generation, editing, and multi-reference synthesis in one sparse model, with strong reported benchmark results and a fast Turbo variant.