A workflow uses the latest Open Source MiniMax H3 to generate videos from multiple references. Other model details are within the workflow.<h3 id="try-the-workflow-online(faster):-basic:https:www.runninghub.aipost2084116916972908545-dual-sampler:https:www.runninghub.aipost2086790597210746881latent-upscaler:https:www.runninghub.aipost2089678677750923266">Try the workflow online(Faster): Basic: <a target="_blank" rel="ugc" href="https://www.runninghub.ai/post/2084116916972908545/?inviteCode=rh-v1152">https://www.runninghub.ai/post/2084116916972908545 </a> Dual Sampler: <a target="_blank" rel="ugc" href="https://www.runninghub.ai/post/2086790597210746881/?inviteCode=rh-v1152">https://www.runninghub.ai/post/2086790597210746881</a> Latent Upscaler: <a target="_blank" rel="ugc" href="https://www.runninghub.ai/post/2089678677750923266/?inviteCode=rh-v1152">https://www.runninghub.ai/post/2089678677750923266</a></h3><h3 id="please-check-the-video-tutorial-before-you-use-this-workflow:"> Please check the video tutorial before you use this workflow: </h3>Basic: <a target="_blank" rel="ugc" href="https://rumble.com/v7dotpw-seedance-is-dead-minimax-h3-rebornfree-open-source.html">https://rumble.com/v7dotpw-seedance-is-dead-minimax-h3-rebornfree-open-source.html</a> Dual Sampler: <a target="_blank" rel="ugc" href="https://rumble.com/v7e3llc-free-open-source-gods-i-make-the-minimax-h3-recreate-a-classic-60-second-sh.html">https://rumble.com/v7e3llc-free-open-source-gods-i-make-the-minimax-h3-recreate-a-classic-60-second-sh.html</a> Latent Upscaler+Local Prompt: <a target="_blank" rel="ugc" href="https://rumble.com/v7een90-minimax-h3-latent-upscaler-local-prompt-skill-nodeqwen3.8-gguf.html">https://rumble.com/v7een90-minimax-h3-latent-upscaler-local-prompt-skill-nodeqwen3.8-gguf.html</a> Custom Nodes Used In Workflow: Custom Nodes: <a target="_blank" rel="ugc" href="https://github.com/T8mars/comfyui-minimax-h3-audio-T8">https://github.com/T8mars/comfyui-minimax-h3-audio-T8</a> <a target="_blank" rel="ugc" href="https://github.com/ptmaster/ComfyUI-PT_H3ConcatAVLatent">https://github.com/ptmaster/ComfyUI-PT_H3ConcatAVLatent</a> <a target="_blank" rel="ugc" href="https://github.com/WhatDreamsCost/WhatDreamsCost-ComfyUI">https://github.com/WhatDreamsCost/WhatDreamsCost-ComfyUI</a> <a target="_blank" rel="ugc" href="https://github.com/smthemex/ComfyUI_UniBlockSwap">https://github.com/smthemex/ComfyUI_UniBlockSwap</a> <a target="_blank" rel="ugc" href="https://github.com/chflame163/ComfyUI_Qwen_H3_Prompt">https://github.com/chflame163/ComfyUI_Qwen_H3_Prompt</a> <a target="_blank" rel="ugc" href="https://github.com/LBH-123-AI/Comfyui_Minimax_h3_latent_Upscaler">https://github.com/LBH-123-AI/Comfyui_Minimax_h3_latent_Upscaler</a> Skill: <a target="_blank" rel="ugc" href="https://github.com/landon2022/minimax-h3-video-prompt">https://github.com/landon2022/minimax-h3-video-prompt</a>

MiniMax H3 Multi-Refs  Video Gen  Accelerated Version(Native NSFW)

This version uses a latent upscaler in the second-stage sampling, saving the time of repeated decoding and encoding. Fix artifacts in the output video. Plus a fully local prompt skill node(Qwen3.8 27b).

Files with SHA256: 4824ddebd110919fd00a410c72d9c0998f6a6bb06e4aa4cbfaeb5bb9bd4628a3