MiniMax's H3-Image Model Adds Local Hosting Support via vLLM and SGLang

vLLM-Omni and SGLang Diffusion platforms now support MiniMax's H3-Image model for local hosting on NVIDIA and AMD GPUs. The model can produce 2048x2048 images in 0.63 seconds on four NVIDIA B300 GPUs. The open-weight license prohibits use in the United States, the European Union, South Korea, and the United Kingdom.