MiniMax-H3 Anime Line Art Colorization LoRA
gens exquisite anime visuals from input line art videos
huggingface.co/DiffSynth-Stud…
Wildminder
@wildmindai
Physicist, Programmer, Designer
- Oh wow, ComfyUI already supports MiniMax Music 3! Just awesome! Looks like we’re in for a fun weekend. huggingface.co/Comfy-Org/Mini…🎵MiniMax-Music3 Next-Generation Open-Weights Production-Ready & Versatile Music Model huggingface.co/MiniMaxAI/Mini…
- These guys are really cooking! No time to breathe, they’ve already released a music model. MiniMax Music 3.0, 60Gb... production-ready.. - 5-minute songs + arrangements + vocals - 8B global/0.6B local - handles structure + acoustics🎵MiniMax-Music3 Next-Generation Open-Weights Production-Ready & Versatile Music Model huggingface.co/MiniMaxAI/Mini…
- MiDashengLM-Gen - One-pass audio scene synthesis. - Qwen3-1.7B - blends speech, music, ambient sound - 2.79% EN WER, 3.87% ZH CER - variable-length output - 9 languages. xingws.github.io/midashenglm-ge…
- ComfyUI Minimax-H3 FaceRefine for refining small faces. - tracks, crops, regenerates faces at native resolution to fix distance-blurring - YOLOv8 + InsightFace for identity-locked tracking - img2img latent injection + temporal denoise scaling - fits in 12GB VRAM via GGUF.
