HunyuanImage 3.0

HunyuanImage 3.0 is Tencent's open-source multimodal image-generation foundation model that unifies text understanding, image generation, and instruction-driven editing in a single autoregressive Mixture-of-Experts architecture. At 80B total parameters (13B active per token), it is presented as the largest open-source image-generation MoE model available. Source: supabase.