Tag
OliviaRossi 发布了 Minimax-h3_Singularity-x-10Eros-Max_PRISM，使用自定义 PRISM 算法（非 mergekit）合并两个 Minimax-H3 Ref2VA 微调模型，包含谱去噪、极坐标测地线插值和符号冲突调解等步骤。
Licon VBVR 是一个针对 MiniMax-H3 的图像转视频 LoRA，基于重建的 VBVR 数据训练，可减少物体复制、裁剪与穿插问题，并提升角色动作逻辑与空间一致性。
A Hugging Face LoRA for MiniMax-H3 (FL2VA variant) that converts a single photo into a seamless 360° frozen-scene camera orbit returning exactly to the starting frame, enabling cleanly chainable video clips.
Akatz Labs released an experimental MiniMax H3 character-swap LoRA (1,000 training steps) that replaces a selected character in a source video using an image reference while preserving the scene. The authors note background preservation is decent, but motion timing, facial expressions, and hard cuts remain unreliable.
HyperFlow introduces a new MiniMax-H3 variant with open-weight LoRA that enables 3× faster inference through data-free self-distillation while preserving video and stereo audio outputs.
HyperFlow is an open-source 8-step LoRA that uses data-free flow self-distillation to reduce video generation steps from 49 to 8 in MiniMax-H3, achieving significant speedup while maintaining quality.
Draw Things app adds support for the MiniMax H3 model, enabling first-and-last-frame generation and reference-based video generation features.
This repository provides LoRAs for the MiniMax H3 model, designed to run in ComfyUI for video enhancement, such as sharpening videos while maintaining photorealism.
A tweet discussing the uncensored version of Minimax H3 AI model, sharing a video generation prompt template, skills, and workflow to improve video output, with a reference to an open-source project.
VDN-Minimax-H3 is an open-source hybrid-attention model that speeds up video generation with near-lossless quality, featuring fast inference and plug-and-play adapters powered by MiniMax H3.
H3-World transforms the MiniMax-H3 video generator into an interactive world model by enabling language-based temporal control, achieving effective control with minimal fine-tuning.
SlopTV is an AI livestream that uses Minimax H3 on dual 5090 GPUs to generate infinite content from YouTube chat comments.
A developer has integrated MiniMax H3 video generation into TensorSharp, a local inference engine, enabling image-to-video generation and highlighting the convergence of LLM, multimodal, and video inference in a single runtime.
Introduces the cost advantages of using cloud GPUs to run the MiniMax H3 model for AI video generation, and plans to open-source a Codex plugin to simplify the workflow.
A user praises Minimax H3 for its impressive performance in handling text, scenes, dialogues, and music in video creation, demonstrating its capabilities in a 30-second video project.
FastVideo has released an open-source, post-trained version of Minimax H3 that generates video 50x faster, producing 5 seconds of video from 3 seconds of compute for text-to-audio-video generation.
Alibaba PAI releases LoRA checkpoints for accelerating MiniMax-H3 video generation using Parallel Decoding Distillation, enabling efficient inference in 8 steps.
MiniMax H3 AI model demonstrated the ability to generate a pop-up anime music video, showcasing its content creation capabilities.
Discovered a platform that uses Minimax H3 to generate 720p videos at a very low cost. A 10-second video was tested to take only 3 minutes, priced at 0.01 RMB per second.
A neural latent-space upscaler designed to accelerate high-resolution Minimax H3 video generation by upscaling latent representations directly, avoiding expensive decode-encode round-trips.