realrebelai/MiniMax-H3_GGUFs
Summary
Hugging Face repository providing GGUF quantizations of MiniMax-H3 models for use with ComfyUI, including directory structure and links to required VAEs.
View Cached Full Text
Cached at: 08/04/26, 07:35 PM
realrebelai/MiniMax-H3_GGUFs Β· Hugging Face
Source: https://huggingface.co/realrebelai/MiniMax-H3_GGUFs
https://huggingface.co/realrebelai/MiniMax-H3_GGUFs#model-card-minimax-h3-ggufsModel Card: MiniMax-H3 GGUFs
https://huggingface.co/realrebelai/MiniMax-H3_GGUFs#directory-structureDirectory Structure
π ComfyUI/
βββ π models/
β βββ π unet/
β β βββ MiniMax-H3-FL2VA-<Q_>.gguf
β β βββ MiniMax-H3-REF2V-<Q_>.gguf
β βββ π text_encoders/
β β βββ qwen3vl_32b_minimax_h3_Q4_K_M.gguf
β βββ π vae/
β β βββ minimax_h3_audio_vae_fp32.safetensors
β β βββ minimax_h3_video_vae_fp16.safetensors
https://huggingface.co/realrebelai/MiniMax-H3_GGUFs#vaes-can-be-found-here-you-need-both-httpshuggingfacecocomfy-orgminimax-h3treemainvaevaes can be found here, you need both:https://huggingface.co/Comfy-Org/MiniMax-H3/tree/main/vae
https://huggingface.co/realrebelai/MiniMax-H3_GGUFs#licensing-and-permissionsLicensing and Permissions
i have permission under a license agreement with MiniMax.
Similar Articles
Kijai/MiniMax-H3_comfy
A Hugging Face repository hosting MiniMax-H3 models converted for ComfyUI usage, along with a Lightx2v distill LoRA for faster inference.
Unsloth Minimax M3 GGUF
Unsloth is uploading a GGUF quantized version of the MiniMax M3 model to Hugging Face.
Comfy-Org/MiniMax-H3
Comfy-Org repackaged MiniMax-H3 model files for ComfyUI, including diffusion models, text encoders, and VAEs, with workflow templates for text-to-video, image-to-video, and reference-to-video generation.
@aisearchio: Minimax H3 GGUFs are here! Q2 is only 8.49 GB, so you could fit it in lower-end GPUs. https://huggingface.co/realrebelaβ¦
Announcement that GGUF quantizations of MiniMax H3 are available, with the Q2 version being only 8.49 GB for lower-end GPUs.
Kijai/MiniMax-H3-experimental
Kijai's MiniMax-H3-experimental is a work-in-progress model featuring w4a8 quantization and int8_convrot VAE support for ComfyUI, released for testing purposes.