AI BriefingMiniMax-H3-Image: MiniMax H3 Image Model Ready to Use in ComfyUI
Comfy-Org/MiniMax-H3-Image
About the project
A file set repackaging a MiniMax H3-based text-to-image generation model for the ComfyUI environment. It organizes the model files from the original repository to fit ComfyUI's diffusion model, text encoder, and VAE folder structures.
For prompt enhancement, it is recommended to use the qwen3.5_9b_w4a8_gptq model with the Generate Text node. It supports int8 and nvfp4 quantized versions based on qwen3vl_32b as the text encoder, and the VAE is configured in fp16 format.
The license terms differ from the original MiniMax H3. If annual revenue exceeds 1 million dollars or if hosting services are provided to third parties, separate prior written approval must be obtained through the MiniMax platform.
Comfy-Org/MiniMax-H3-Image
The original page has no description.
text-to-image
This introduction was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report errors, attribution issues, or removal requests via Contact.