An open-weight image model for text-to-image, references and editing
Grounded in available product and source data
MiniMax H3-Image is an open-weight image generation model built on H3. It handles text-to-image work, reference-based generation and image editing, and the team has published the model weights so developers can adapt it.
The published results cover speed, image quality and prompt adherence at the same time. The model is also available through the MiniMax API Platform, MiniMax Code and Hailuo AI.