{"id":"b218a02d-0837-4257-98c0-06582e77346d","revision":0,"last_node_id":156,"last_link_id":300,"nodes":[{"id":92,"type":"SaveVideo","pos":[1150,4880],"size":[1210,106],"flags":{},"order":34,"mode":0,"inputs":[{"localized_name":"视频","name":"video","type":"VIDEO","link":260},{"localized_name":"文件名前缀","name":"filename_prefix","type":"STRING","widget":{"name":"filename_prefix"},"link":null},{"localized_name":"格式","name":"format","type":"COMBO","widget":{"name":"format"},"link":null},{"localized_name":"编码器","name":"codec","type":"COMFY_DYNAMICCOMBO_V3","widget":{"name":"codec"},"link":null}],"outputs":[{"localized_name":"video","name":"video","type":"VIDEO","links":null}],"properties":{},"widgets_values":["video/MiniMax_H3","auto","auto"]},{"id":116,"type":"MarkdownNote","pos":[-2030,4850],"size":[450,889.6875],"flags":{},"order":0,"mode":0,"inputs":[],"outputs":[],"title":"Note: MiniMax H3","properties":{},"widgets_values":["## MiniMax H3\n\n[MiniMax H3](https://www.minimax.io/blog/minimax-h3) is MiniMax's general-purpose, omni-modal generation model. It jointly understands text, image, video, and audio, and generates video with **native stereo audio**: voice, sound effects, and music are modeled jointly in a single forward pass, not layered on afterward. Output is up to 2K resolution, 24fps, and up to about 15 seconds.\n\n## ComfyUI links\n- [ComfyUI#15224](https://github.com/Comfy-Org/ComfyUI/pull/15224)\n- [🤗 Comfy-Org/MiniMax-H3](https://huggingface.co/Comfy-Org/MiniMax-H3)\n\n## About this workflow\n\nThis template runs the **reference-to-video (ref2va)** task using the `MiniMaxH3ReferenceToVideo` node. It takes any mix of reference images, videos, and standalone audio, and weaves them into the generation to lock in a character's identity, a style, a motion, a camera move, or a voice.\n\n**Key inputs**\n\n- **ref_images / ref_videos / ref_video_audios / ref_audios**: up to 9 reference images, 3 reference videos (each may carry its own paired soundtrack), and 3 standalone reference audio clips\n- **prompt**: reference the inputs by tag, in the exact order they were connected, for example `<Picture 1>`, `<Video 1>`, `<Audio 1>`, then describe the target scene, motion, and audio\n- **ref_image_size**: `match` scales references down to the generation's resolution (faster); `max` keeps up to a 2048px short edge for stronger identity fidelity, at the cost of speed since reference tokens ride along every sampling step\n- **width / height**: set via Resolution Selector.\n- **duration (seconds)**: converted to a valid frame `length` by the Math Expression node\n\n**Sampling and decode**\n\n- Sampler: `res_multistep`. `beta` or `normal` scheduler tends to outperform `simple` for reference-heavy prompts like this one\n- The sampler's joint audio+video `LATENT` output feeds directly into both `VAEDecode` (video, `minimax_h3_video_vae_fp16`) and `VAEDecodeAudio` (audio, `minimax_h3_audio_vae_fp32`); each decode node automatically pulls its own half out of the packed latent. `CreateVideo` then muxes the two into a single MP4 with synced sound\n- The diffusion model here is `minimax_h3_ref2va_pruned_int8_convrot.safetensors`, a different set of weights from the `fl2va` model used by the t2v/i2v templates\n\nRef2va's output is very sensitive to prompt wording; matching the reference tags precisely and being explicit about which reference drives which part of the shot tends to work best."],"color":"#222","bgcolor":"#000"},{"id":117,"type":"MarkdownNote","pos":[-2500,4850],"size":[440,740],"flags":{},"order":1,"mode":0,"inputs":[],"outputs":[],"title":"Note: Model Links","properties":{},"widgets_values":["## Model Links\n[🤗 Comfy-Org/MiniMax-H3](https://huggingface.co/Comfy-Org/MiniMax-H3)\n\n**vae**\n\n- [minimax_h3_video_vae_fp16.safetensors](https://huggingface.co/Comfy-Org/MiniMax-H3/resolve/main/vae/minimax_h3_video_vae_fp16.safetensors)\n- [minimax_h3_audio_vae_fp32.safetensors](https://huggingface.co/Comfy-Org/MiniMax-H3/resolve/main/vae/minimax_h3_audio_vae_fp32.safetensors)\n\n**diffusion_models**\n\n- [minimax_h3_ref2va_pruned_int8_convrot.safetensors](https://huggingface.co/Comfy-Org/MiniMax-H3/resolve/main/diffusion_models/minimax_h3_ref2va_pruned_int8_convrot.safetensors)\n\n**text_encoders**\n\n- [qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors](https://huggingface.co/Comfy-Org/MiniMax-H3/resolve/main/text_encoders/qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors)\n\n\n## Model Storage Location\n\n```\n📂 ComfyUI/\n├── 📂 models/\n│   ├── 📂 vae/\n│   │   ├── minimax_h3_video_vae_fp16.safetensors\n│   │   └── minimax_h3_audio_vae_fp32.safetensors\n│   ├── 📂 diffusion_models/\n│   │   └── minimax_h3_ref2va_pruned_int8_convrot.safetensors\n│   └── 📂 text_encoders/\n│       └── qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors\n```\n\n## Report Issue\n\nNote: Please update ComfyUI first ([guide](https://docs.comfy.org/installation/update_comfyui)) and prepare required models. Desktop/Cloud updates follow stable releases, so some nightly-supported models may not be available yet.\n\n- Cannot run / runtime errors: [ComfyUI/issues](https://github.com/comfyanonymous/ComfyUI/issues)\n- UI / frontend issues: [ComfyUI_frontend/issues](https://github.com/Comfy-Org/ComfyUI_frontend/issues)\n- Workflow issues: [workflow_templates/issues](https://github.com/Comfy-Org/workflow_templates/issues)\n"],"color":"#222","bgcolor":"#000"},{"id":119,"type":"VAELoader","pos":[-1490,5270],"size":[640,70],"flags":{},"order":2,"mode":0,"inputs":[{"localized_name":"vae名称","name":"vae_name","type":"COMBO","widget":{"name":"vae_name"},"link":null}],"outputs":[{"localized_name":"VAE","name":"VAE","type":"VAE","links":[251,273]}],"properties":{"Node name for S&R":"VAELoader","models":[{"name":"minimax_h3_video_vae_fp16.safetensors","url":"https://huggingface.co/Comfy-Org/MiniMax-H3/resolve/main/vae/minimax_h3_video_vae_fp16.safetensors","directory":"vae"}]},"widgets_values":["minimax_h3_video_vae_fp16.safetensors"]},{"id":120,"type":"VAELoader","pos":[-1490,5400],"size":[650,70],"flags":{},"order":3,"mode":0,"inputs":[{"localized_name":"vae名称","name":"vae_name","type":"COMBO","widget":{"name":"vae_name"},"link":null}],"outputs":[{"localized_name":"VAE","name":"VAE","type":"VAE","links":[250,274]}],"properties":{"Node name for S&R":"VAELoader","models":[{"name":"minimax_h3_audio_vae_fp32.safetensors","url":"https://huggingface.co/Comfy-Org/MiniMax-H3/resolve/main/vae/minimax_h3_audio_vae_fp32.safetensors","directory":"vae"}]},"widgets_values":["minimax_h3_audio_vae_fp32.safetensors"]},{"id":123,"type":"KSamplerSelect","pos":[-260,5190],"size":[370,70],"flags":{},"order":4,"mode":0,"inputs":[{"localized_name":"采样器名称","name":"sampler_name","type":"COMBO","widget":{"name":"sampler_name"},"link":null}],"outputs":[{"localized_name":"采样器","name":"SAMPLER","type":"SAMPLER","links":[255]}],"properties":{"Node name for S&R":"KSamplerSelect"},"widgets_values":["res_multistep"]},{"id":129,"type":"RandomNoise","pos":[-260,4940],"size":[360,90],"flags":{},"order":5,"mode":0,"inputs":[{"localized_name":"噪波随机种","name":"noise_seed","type":"INT","widget":{"name":"noise_seed"},"link":null}],"outputs":[{"localized_name":"噪波","name":"NOISE","type":"NOISE","links":[253]}],"properties":{"Node name for S&R":"RandomNoise"},"widgets_values":[79465993831824,"randomize"]},{"id":131,"type":"ComfyMathExpression","pos":[-1450,5940],"size":[225,74],"flags":{"collapsed":true},"order":23,"mode":0,"inputs":[{"label":"a","localized_name":"values.a","name":"values.a","type":"FLOAT,INT,BOOLEAN","link":261},{"label":"b","localized_name":"values.b","name":"values.b","shape":7,"type":"FLOAT,INT,BOOLEAN","link":null},{"localized_name":"表达式","name":"expression","type":"STRING","widget":{"name":"expression"},"link":null}],"outputs":[{"localized_name":"浮点","name":"FLOAT","type":"FLOAT","links":[]},{"localized_name":"整数","name":"INT","type":"INT","links":[275]},{"localized_name":"布尔值","name":"BOOL","type":"BOOLEAN","links":[]}],"properties":{"Node name for S&R":"ComfyMathExpression"},"widgets_values":["max(5, round(a * 24)) + (5 - (max(5, round(a * 24)) % 17)) % 17"]},{"id":140,"type":"MarkdownNote","pos":[-1890,5830],"size":[300,520],"flags":{},"order":6,"mode":0,"inputs":[],"outputs":[],"title":"Note: Size Settings Reference","properties":{},"widgets_values":["| megapixels | Aspect | Output (multiple=32) |\n|---|---|---|\n| 0.2 | 16:9 | 608 x 352 |\n| 0.3 | 16:9 | 736 x 416 |\n| 0.4 | 16:9 | 864 x 480 |\n| 0.5 | 16:9 | 960 x 544 |\n| 0.6 | 16:9 | 1056 x 608 |\n| 0.7 | 16:9 | 1152 x 640 |\n| 0.8 | 16:9 | 1216 x 672 |\n| 0.9 | 16:9 | 1280 x 736 |\n| 0.98 | 16:9 | 1344 x 768 |\n| 1.0 | 16:9 | 1376 x 768 |\n| 1.2 | 16:9 | 1504 x 832 |\n| 1.5 | 16:9 | 1664 x 928 |\n| 1.8 | 16:9 | 1824 x 1024 |\n| 2.0 | 16:9 | 1920 x 1088 |\n"],"color":"#222","bgcolor":"#000"},{"id":130,"type":"CreateVideo","pos":[790,5090],"size":[270,110],"flags":{},"order":33,"mode":0,"inputs":[{"localized_name":"图像","name":"images","type":"IMAGE","link":286},{"localized_name":"音频","name":"audio","shape":7,"type":"AUDIO","link":259},{"localized_name":"帧率","name":"fps","type":"FLOAT","widget":{"name":"fps"},"link":null},{"localized_name":"bit_depth","name":"bit_depth","shape":7,"type":"INT","widget":{"name":"bit_depth"},"link":null}],"outputs":[{"localized_name":"视频","name":"VIDEO","type":"VIDEO","links":[260]}],"properties":{"Node name for S&R":"CreateVideo"},"widgets_values":[24,8]},{"id":144,"type":"LoadAudio","pos":[-864.6372699170003,6056.852994682291],"size":[270,136],"flags":{},"order":7,"mode":4,"inputs":[{"localized_name":"音频","name":"audio","type":"COMBO","widget":{"name":"audio"},"link":null},{"localized_name":"音频UI","name":"audioUI","type":"AUDIO_UI","widget":{"name":"audioUI"},"link":null},{"localized_name":"选择文件上传","name":"upload","type":"AUDIOUPLOAD","widget":{"name":"upload"},"link":null}],"outputs":[{"localized_name":"音频","name":"AUDIO","type":"AUDIO","links":[284]}],"properties":{"Node name for S&R":"LoadAudio"},"widgets_values":["xxt_mv2_S3.wav",null,null]},{"id":150,"type":"MiniMaxH3TurboLoRA","pos":[-984.8000489744444,4628.836792893106],"size":[270,106],"flags":{},"order":21,"mode":0,"inputs":[{"localized_name":"model","name":"model","type":"MODEL","link":291},{"localized_name":"lora_name","name":"lora_name","type":"COMBO","widget":{"name":"lora_name"},"link":null},{"localized_name":"strength","name":"strength","type":"FLOAT","widget":{"name":"strength"},"link":null},{"localized_name":"low_vram","name":"low_vram","type":"BOOLEAN","widget":{"name":"low_vram"},"link":null}],"outputs":[{"localized_name":"模型","name":"MODEL","type":"MODEL","links":[290]}],"properties":{"Node name for S&R":"MiniMaxH3TurboLoRA"},"widgets_values":["minimax_h3_turbo_v4_step600_ema.safetensors",1,false]},{"id":126,"type":"BasicGuider","pos":[-260,5080],"size":[360,60],"flags":{},"order":27,"mode":0,"inputs":[{"localized_name":"模型","name":"model","type":"MODEL","link":292},{"localized_name":"条件","name":"conditioning","type":"CONDITIONING","link":270}],"outputs":[{"localized_name":"引导器","name":"GUIDER","type":"GUIDER","links":[254]}],"properties":{"Node name for S&R":"BasicGuider"},"widgets_values":[]},{"id":128,"type":"CLIPLoader","pos":[-1490,5090],"size":[640,120],"flags":{},"order":8,"mode":0,"inputs":[{"localized_name":"CLIP名称","name":"clip_name","type":"COMBO","widget":{"name":"clip_name"},"link":null},{"localized_name":"类型","name":"type","type":"COMBO","widget":{"name":"type"},"link":null},{"localized_name":"设备","name":"device","shape":7,"type":"COMBO","widget":{"name":"device"},"link":null}],"outputs":[{"localized_name":"CLIP","name":"CLIP","type":"CLIP","links":[272]}],"properties":{"Node name for S&R":"CLIPLoader","models":[{"name":"qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors","url":"https://huggingface.co/Comfy-Org/MiniMax-H3/resolve/main/text_encoders/qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors","directory":"text_encoders"}]},"widgets_values":["qwen3vl_32b_h3_ultra_uncensored_heretic_int8_convrot.safetensors","minimax","default"]},{"id":127,"type":"UNETLoader","pos":[-1490,4940],"size":[640,90],"flags":{},"order":9,"mode":0,"inputs":[{"localized_name":"UNet名称","name":"unet_name","type":"COMBO","widget":{"name":"unet_name"},"link":null},{"localized_name":"数据类型","name":"weight_dtype","type":"COMBO","widget":{"name":"weight_dtype"},"link":null}],"outputs":[{"localized_name":"模型","name":"MODEL","type":"MODEL","links":[291]}],"properties":{"Node name for S&R":"UNETLoader","models":[{"name":"minimax_h3_ref2va_pruned_int8_convrot.safetensors","url":"https://huggingface.co/Comfy-Org/MiniMax-H3/resolve/main/diffusion_models/minimax_h3_ref2va_pruned_int8_convrot.safetensors","directory":"diffusion_models"}]},"widgets_values":["minimax_h3_ref2va_pruned_fp8_scaled.safetensors","default"]},{"id":147,"type":"RTXVideoSuperResolution","pos":[808.2209533541876,4901.564327805565],"size":[285.81015625,106],"flags":{},"order":32,"mode":0,"inputs":[{"localized_name":"images","name":"images","type":"IMAGE","link":285},{"localized_name":"resize_type","name":"resize_type","type":"COMFY_DYNAMICCOMBO_V3","widget":{"name":"resize_type"},"link":null},{"localized_name":"resize_type.scale","name":"resize_type.scale","type":"FLOAT","widget":{"name":"resize_type.scale"},"link":null},{"localized_name":"quality","name":"quality","type":"COMBO","widget":{"name":"quality"},"link":null}],"outputs":[{"localized_name":"upscaled_images","name":"upscaled_images","type":"IMAGE","links":[286]}],"properties":{"Node name for S&R":"RTXVideoSuperResolution"},"widgets_values":["scale by multiplier",2,"ULTRA"]},{"id":124,"type":"BasicScheduler","pos":[-260,5310],"size":[370,130],"flags":{},"order":26,"mode":0,"inputs":[{"localized_name":"模型","name":"model","type":"MODEL","link":293},{"localized_name":"调度器","name":"scheduler","type":"COMBO","widget":{"name":"scheduler"},"link":null},{"localized_name":"步数","name":"steps","type":"INT","widget":{"name":"steps"},"link":null},{"localized_name":"降噪","name":"denoise","type":"FLOAT","widget":{"name":"denoise"},"link":null}],"outputs":[{"localized_name":"Sigmas","name":"SIGMAS","type":"SIGMAS","links":[256]}],"properties":{"Node name for S&R":"BasicScheduler"},"widgets_values":["simple",8,1]},{"id":154,"type":"LoadImage","pos":[839.9879664914482,5623.887623374108],"size":[290,330],"flags":{},"order":10,"mode":4,"inputs":[{"localized_name":"图像","name":"image","type":"COMBO","widget":{"name":"image"},"link":null},{"localized_name":"选择文件上传","name":"upload","type":"IMAGEUPLOAD","widget":{"name":"upload"},"link":null}],"outputs":[{"localized_name":"图像","name":"IMAGE","type":"IMAGE","links":[296]},{"localized_name":"遮罩","name":"MASK","type":"MASK","links":null}],"properties":{"Node name for S&R":"LoadImage"},"widgets_values":["WhatsApp Image 2026-08-14 at 3.41.08 AM (1).jpeg","image"]},{"id":122,"type":"VAEDecode","pos":[503.8070945141654,4933.028829257785],"size":[230,60],"flags":{"collapsed":false},"order":31,"mode":0,"inputs":[{"localized_name":"Latent","name":"samples","type":"LATENT","link":300},{"localized_name":"vae","name":"vae","type":"VAE","link":251}],"outputs":[{"localized_name":"图像","name":"IMAGE","type":"IMAGE","links":[285]}],"properties":{"Node name for S&R":"VAEDecode"},"widgets_values":[]},{"id":121,"type":"VAEDecodeAudio","pos":[498.1387386774594,5057.563513927154],"size":[230,60],"flags":{"collapsed":false},"order":30,"mode":0,"inputs":[{"localized_name":"Latent","name":"samples","type":"LATENT","link":299},{"localized_name":"vae","name":"vae","type":"VAE","link":250}],"outputs":[{"localized_name":"音频","name":"AUDIO","type":"AUDIO","links":[259]}],"properties":{"Node name for S&R":"VAEDecodeAudio"},"widgets_values":[]},{"id":156,"type":"VRAM_Debug","pos":[432.7511439303426,5275.471471196973],"size":[304.5,186],"flags":{},"order":29,"mode":0,"inputs":[{"localized_name":"any_input","name":"any_input","shape":7,"type":"*","link":298},{"localized_name":"image_pass","name":"image_pass","shape":7,"type":"IMAGE","link":null},{"localized_name":"model_pass","name":"model_pass","shape":7,"type":"MODEL","link":null},{"localized_name":"empty_cache","name":"empty_cache","type":"BOOLEAN","widget":{"name":"empty_cache"},"link":null},{"localized_name":"gc_collect","name":"gc_collect","type":"BOOLEAN","widget":{"name":"gc_collect"},"link":null},{"localized_name":"unload_all_models","name":"unload_all_models","type":"BOOLEAN","widget":{"name":"unload_all_models"},"link":null}],"outputs":[{"localized_name":"any_output","name":"any_output","type":"*","links":[299,300]},{"localized_name":"image_pass","name":"image_pass","type":"IMAGE","links":null},{"localized_name":"model_pass","name":"model_pass","type":"MODEL","links":null},{"label":"32,142,168,688   freemem_before","localized_name":"freemem_before","name":"freemem_before","type":"INT","links":null},{"label":"32,145,584,380      freemem_after","localized_name":"freemem_after","name":"freemem_after","type":"INT","links":null}],"properties":{"Node name for S&R":"VRAM_Debug"},"widgets_values":[true,true,true]},{"id":153,"type":"LoadImage","pos":[469.4180699956278,5613.3875610221985],"size":[290,330],"flags":{},"order":11,"mode":4,"inputs":[{"localized_name":"图像","name":"image","type":"COMBO","widget":{"name":"image"},"link":null},{"localized_name":"选择文件上传","name":"upload","type":"IMAGEUPLOAD","widget":{"name":"upload"},"link":null}],"outputs":[{"localized_name":"图像","name":"IMAGE","type":"IMAGE","links":[295]},{"localized_name":"遮罩","name":"MASK","type":"MASK","links":null}],"properties":{"Node name for S&R":"LoadImage"},"widgets_values":["WhatsApp Image 2026-08-14 at 3.41.08 AM.jpeg","image"]},{"id":155,"type":"LoadImage","pos":[846.9838512651106,6053.467371472379],"size":[290,330],"flags":{},"order":12,"mode":4,"inputs":[{"localized_name":"图像","name":"image","type":"COMBO","widget":{"name":"image"},"link":null},{"localized_name":"选择文件上传","name":"upload","type":"IMAGEUPLOAD","widget":{"name":"upload"},"link":null}],"outputs":[{"localized_name":"图像","name":"IMAGE","type":"IMAGE","links":[297]},{"localized_name":"遮罩","name":"MASK","type":"MASK","links":null}],"properties":{"Node name for S&R":"LoadImage"},"widgets_values":["WhatsApp Image 2026-08-14 at 3.41.08 AM (2).jpeg","image"]},{"id":149,"type":"LoadVideo","pos":[-540.0075321112383,6079.614440703327],"size":[270,225.875],"flags":{},"order":13,"mode":4,"inputs":[{"localized_name":"文件","name":"file","type":"COMBO","widget":{"name":"file"},"link":null},{"localized_name":"选择要上传的文件","name":"upload","type":"IMAGEUPLOAD","widget":{"name":"upload"},"link":null}],"outputs":[{"localized_name":"视频","name":"VIDEO","type":"VIDEO","links":[287]}],"properties":{"Node name for S&R":"LoadVideo"},"widgets_values":["gp7yg5.mp4","image"]},{"id":148,"type":"GetVideoComponents","pos":[-230.74203766056058,6047.191445317372],"size":[140,86],"flags":{},"order":22,"mode":4,"inputs":[{"localized_name":"视频","name":"video","type":"VIDEO","link":287}],"outputs":[{"localized_name":"图像","name":"images","type":"IMAGE","links":[288]},{"localized_name":"音频","name":"audio","type":"AUDIO","links":[289]},{"localized_name":"帧率","name":"fps","type":"FLOAT","links":null},{"localized_name":"bit_depth","name":"bit_depth","type":"INT","links":null}],"properties":{"Node name for S&R":"GetVideoComponents"},"widgets_values":[]},{"id":136,"type":"MiniMaxH3ReferenceToVideo","pos":[-751.1336711673416,4929.7969594939295],"size":[410,530],"flags":{},"order":25,"mode":0,"inputs":[{"localized_name":"clip","name":"clip","type":"CLIP","link":272},{"localized_name":"vae","name":"vae","type":"VAE","link":273},{"localized_name":"audio_vae","name":"audio_vae","type":"VAE","link":274},{"label":"ref_image_0","localized_name":"ref_images.ref_image_0","name":"ref_images.ref_image_0","shape":7,"type":"IMAGE","link":278},{"label":"ref_image_1","localized_name":"ref_images.ref_image_1","name":"ref_images.ref_image_1","shape":7,"type":"IMAGE","link":282},{"label":"ref_image_2","localized_name":"ref_images.ref_image_2","name":"ref_images.ref_image_2","shape":7,"type":"IMAGE","link":283},{"label":"ref_image_3","localized_name":"ref_images.ref_image_3","name":"ref_images.ref_image_3","shape":7,"type":"IMAGE","link":294},{"label":"ref_image_4","localized_name":"ref_images.ref_image_4","name":"ref_images.ref_image_4","shape":7,"type":"IMAGE","link":295},{"label":"ref_image_5","localized_name":"ref_images.ref_image_5","name":"ref_images.ref_image_5","shape":7,"type":"IMAGE","link":296},{"label":"ref_image_6","localized_name":"ref_images.ref_image_6","name":"ref_images.ref_image_6","shape":7,"type":"IMAGE","link":297},{"label":"ref_image_7","localized_name":"ref_images.ref_image_7","name":"ref_images.ref_image_7","shape":7,"type":"IMAGE","link":null},{"label":"ref_video_0","localized_name":"ref_videos.ref_video_0","name":"ref_videos.ref_video_0","shape":7,"type":"IMAGE","link":288},{"label":"ref_video_1","localized_name":"ref_videos.ref_video_1","name":"ref_videos.ref_video_1","shape":7,"type":"IMAGE","link":null},{"label":"ref_video_audio_0","localized_name":"ref_video_audios.ref_video_audio_0","name":"ref_video_audios.ref_video_audio_0","shape":7,"type":"AUDIO","link":289},{"label":"ref_video_audio_1","localized_name":"ref_video_audios.ref_video_audio_1","name":"ref_video_audios.ref_video_audio_1","shape":7,"type":"AUDIO","link":null},{"label":"ref_audio_0","localized_name":"ref_audios.ref_audio_0","name":"ref_audios.ref_audio_0","shape":7,"type":"AUDIO","link":284},{"label":"ref_audio_1","localized_name":"ref_audios.ref_audio_1","name":"ref_audios.ref_audio_1","shape":7,"type":"AUDIO","link":null},{"localized_name":"prompt","name":"prompt","type":"STRING","widget":{"name":"prompt"},"link":279},{"localized_name":"width","name":"width","type":"INT","widget":{"name":"width"},"link":276},{"localized_name":"height","name":"height","type":"INT","widget":{"name":"height"},"link":277},{"localized_name":"length","name":"length","type":"INT","widget":{"name":"length"},"link":275},{"localized_name":"ref_image_size","name":"ref_image_size","type":"COMBO","widget":{"name":"ref_image_size"},"link":null}],"outputs":[{"localized_name":"positive","name":"positive","type":"CONDITIONING","links":[270]},{"localized_name":"Latent","name":"LATENT","type":"LATENT","links":[271]}],"properties":{"Node name for S&R":"MiniMaxH3ReferenceToVideo"},"widgets_values":["",1344,768,124,"match"]},{"id":125,"type":"SamplerCustomAdvanced","pos":[150,5110],"size":[230,326],"flags":{},"order":28,"mode":0,"inputs":[{"localized_name":"噪波","name":"noise","type":"NOISE","link":253},{"localized_name":"引导器","name":"guider","type":"GUIDER","link":254},{"localized_name":"采样器","name":"sampler","type":"SAMPLER","link":255},{"localized_name":"西格玛","name":"sigmas","type":"SIGMAS","link":256},{"localized_name":"Latent图像","name":"latent_image","type":"LATENT","link":271}],"outputs":[{"localized_name":"Latent","name":"output","type":"LATENT","links":[298]},{"localized_name":"降噪Latent","name":"denoised_output","type":"LATENT","links":[]}],"properties":{"Node name for S&R":"SamplerCustomAdvanced"},"widgets_values":[]},{"id":132,"type":"PrimitiveFloat","pos":[-1480,5830],"size":[270,70],"flags":{},"order":14,"mode":0,"inputs":[{"localized_name":"值","name":"value","type":"FLOAT","widget":{"name":"value"},"link":null}],"outputs":[{"localized_name":"浮点","name":"FLOAT","type":"FLOAT","links":[261]}],"title":"Float (Duration)","properties":{"Node name for S&R":"PrimitiveFloat"},"widgets_values":[12],"color":"#322","bgcolor":"#533"},{"id":151,"type":"MiniMaxH3MemoryEfficientSolAttentionPatch","pos":[-517.5728416466122,4525.8261390551925],"size":[430,250],"flags":{},"order":24,"mode":0,"inputs":[{"localized_name":"model","name":"model","type":"MODEL","link":290},{"localized_name":"enabled","name":"enabled","type":"BOOLEAN","widget":{"name":"enabled"},"link":null},{"localized_name":"tau","name":"tau","type":"FLOAT","widget":{"name":"tau"},"link":null},{"localized_name":"min_tokens","name":"min_tokens","type":"INT","widget":{"name":"min_tokens"},"link":null},{"localized_name":"strict","name":"strict","type":"BOOLEAN","widget":{"name":"strict"},"link":null},{"localized_name":"thresh_type","name":"thresh_type","type":"COMBO","widget":{"name":"thresh_type"},"link":null},{"localized_name":"int8_qk","name":"int8_qk","type":"BOOLEAN","widget":{"name":"int8_qk"},"link":null},{"localized_name":"int8_pv","name":"int8_pv","type":"BOOLEAN","widget":{"name":"int8_pv"},"link":null},{"localized_name":"sink_conditioning","name":"sink_conditioning","type":"COMBO","widget":{"name":"sink_conditioning"},"link":null},{"localized_name":"dense_blocks","name":"dense_blocks","type":"STRING","widget":{"name":"dense_blocks"},"link":null}],"outputs":[{"localized_name":"模型","name":"MODEL","type":"MODEL","links":[292,293]}],"properties":{"Node name for S&R":"MiniMaxH3MemoryEfficientSolAttentionPatch","aux_id":"Saganaki22/ComfyUI-sol-attn","ver":"e2fc225f8642585cfa11a31d52fe7b2db7290efa","ue_properties":{"widget_ue_connectable":{},"input_ue_unconnectable":{},"version":"7.8"}},"widgets_values":[true,1,4096,true,"diag",false,"exact_kv","exact_kv",""]},{"id":137,"type":"LoadImage","pos":[-896.4583891150304,5637.486243506594],"size":[290,330],"flags":{},"order":15,"mode":0,"inputs":[{"localized_name":"图像","name":"image","type":"COMBO","widget":{"name":"image"},"link":null},{"localized_name":"选择文件上传","name":"upload","type":"IMAGEUPLOAD","widget":{"name":"upload"},"link":null}],"outputs":[{"localized_name":"图像","name":"IMAGE","type":"IMAGE","links":[278]},{"localized_name":"遮罩","name":"MASK","type":"MASK","links":null}],"title":"加载图像1","properties":{"Node name for S&R":"LoadImage"},"widgets_values":["h3_product_d4b54e4d_1_d54ec0b5.jpg","image"]},{"id":141,"type":"LoadImage","pos":[-221.74481186532637,5630.333359720593],"size":[290,330],"flags":{},"order":16,"mode":0,"inputs":[{"localized_name":"图像","name":"image","type":"COMBO","widget":{"name":"image"},"link":null},{"localized_name":"选择文件上传","name":"upload","type":"IMAGEUPLOAD","widget":{"name":"upload"},"link":null}],"outputs":[{"localized_name":"图像","name":"IMAGE","type":"IMAGE","links":[283]},{"localized_name":"遮罩","name":"MASK","type":"MASK","links":null}],"title":"加载图像3","properties":{"Node name for S&R":"LoadImage"},"widgets_values":["h3_product_b0d902a4_3_0ca5b04f.jpg","image"]},{"id":152,"type":"LoadImage","pos":[121.93169041721814,5642.407213548718],"size":[290,330],"flags":{},"order":17,"mode":0,"inputs":[{"localized_name":"图像","name":"image","type":"COMBO","widget":{"name":"image"},"link":null},{"localized_name":"选择文件上传","name":"upload","type":"IMAGEUPLOAD","widget":{"name":"upload"},"link":null}],"outputs":[{"localized_name":"图像","name":"IMAGE","type":"IMAGE","links":[294]},{"localized_name":"遮罩","name":"MASK","type":"MASK","links":null}],"title":"加载图像4","properties":{"Node name for S&R":"LoadImage"},"widgets_values":["h3_product_b0d902a4_4_a2a94615.jpg","image"]},{"id":115,"type":"ResolutionSelector","pos":[-1480,5610],"size":[270,170],"flags":{},"order":18,"mode":0,"showAdvanced":false,"inputs":[{"localized_name":"宽高比","name":"aspect_ratio","type":"COMBO","widget":{"name":"aspect_ratio"},"link":null},{"localized_name":"百万像素","name":"megapixels","type":"FLOAT","widget":{"name":"megapixels"},"link":null},{"localized_name":"倍数","name":"multiple","type":"INT","widget":{"name":"multiple"},"link":null}],"outputs":[{"localized_name":"宽度","name":"width","type":"INT","links":[276]},{"localized_name":"高度","name":"height","type":"INT","links":[277]}],"title":"Resolution Selector (Size)","properties":{"Node name for S&R":"ResolutionSelector"},"widgets_values":["16:9 (Widescreen)",1,32],"color":"#322","bgcolor":"#533"},{"id":139,"type":"LoadImage","pos":[-546.4247838120406,5624.426010929818],"size":[290,330],"flags":{},"order":19,"mode":0,"inputs":[{"localized_name":"图像","name":"image","type":"COMBO","widget":{"name":"image"},"link":null},{"localized_name":"选择文件上传","name":"upload","type":"IMAGEUPLOAD","widget":{"name":"upload"},"link":null}],"outputs":[{"localized_name":"图像","name":"IMAGE","type":"IMAGE","links":[282]},{"localized_name":"遮罩","name":"MASK","type":"MASK","links":null}],"title":"加载图像2","properties":{"Node name for S&R":"LoadImage"},"widgets_values":["h3_product_61d06cdf_2_6324c4b9.jpg","image"]},{"id":138,"type":"PrimitiveStringMultiline","pos":[-1480,6000],"size":[370,390],"flags":{},"order":20,"mode":0,"inputs":[{"localized_name":"值","name":"value","type":"STRING","widget":{"name":"value"},"link":null}],"outputs":[{"localized_name":"字符串","name":"STRING","type":"STRING","links":[279]}],"title":"Input Text (Prompt)","properties":{"Node name for S&R":"PrimitiveStringMultiline"},"widgets_values":["Create one continuous 10-second photorealistic commercial product demonstration video using the supplied reference images.\n\nREFERENCE ROLES\n\n<Picture 1> is the identity reference for the main model.\nPreserve the model’s facial identity, hairstyle, apparent age, body proportions, skin appearance, and recognizable personal features consistently throughout the entire video.\n\n<Picture 2> is Product A.\nExtract only the actual product from this reference. Preserve its original shape, proportions, construction, materials, surface texture, color relationships, hardware, decorations, functional details, and recognizable design.\n\n<Picture 3> is Product B.\nExtract only the actual product from this reference. Preserve its original shape, proportions, construction, materials, surface texture, color relationships, patterns, seams, accessories, and recognizable design.\n\n<Picture 4> is the environment reference.\nReconstruct the location, spatial layout, environmental scale, surfaces, lighting conditions, background elements, depth, atmosphere, and overall visual character shown in this reference.\n\nAUTOMATIC PRODUCT AND ENVIRONMENT ADAPTATION\n\nFirst, visually analyze Product A and Product B and automatically determine their correct categories, scale, physical properties, intended uses, attachment methods, and natural relationships with the human body.\n\nIntegrate both products naturally with the model according to their actual functions:\n\n- If a product is clothing, the model wears it correctly with realistic fit, folds, fabric tension, stitching, layering, and movement.\n- If a product is footwear, the model wears it correctly on the appropriate feet.\n- If a product is a handbag, backpack, or carryable object, the model naturally carries, holds, wears, opens, adjusts, or presents it according to its real design.\n- If a product is jewelry, eyewear, a watch, a hat, or another accessory, place it correctly on the appropriate part of the model’s body.\n- If a product is handheld, the model holds and demonstrates it with believable hand placement and finger contact.\n- If a product is functional equipment, the model uses it according to its intended purpose.\n- If a product cannot reasonably be worn, attached, or carried, position it immediately beside the model and let the model interact with it naturally while keeping the complete model visible.\n- Never force a product into an anatomically or functionally incorrect position.\n\nAutomatically interpret the environment in <Picture 4> and create actions appropriate to that location. The model’s walking style, poses, gestures, product handling, and environmental interaction must feel naturally motivated by the supplied scene rather than following a fixed activity.\n\nThe products from <Picture 2> and <Picture 3> must be visibly integrated with the model, not presented as isolated floating objects or separate catalogue images.\n\nPRIMARY FULL-BODY COMPOSITION REQUIREMENT\n\nThe complete main model is the highest visual priority throughout the entire video.\n\nMaintain a continuous head-to-toe full-body composition in every stage of the 10-second video. The model’s complete head, hair, face, torso, arms, hands, hips, legs, and both feet must remain visible inside the frame, with comfortable empty space above the hair and below both feet.\n\nKeep the model approximately 65–80 percent of the frame height. Use a wide full-body fashion framing that allows the viewer to clearly understand how Product A and Product B look when worn, carried, attached, held, or used by the model.\n\nProduct details must be revealed through the model’s poses, body rotation, hand gestures, walking direction, and camera angle. Do not obtain product details by cropping the person or moving into an isolated close-up.\n\nIf the system must choose between showing a small product detail and preserving the complete person, always preserve the complete head-to-toe person.\n\nVIDEO STRUCTURE AND TIMING\n\n0.0–2.0 seconds — Full-body introduction:\nBegin with a stable head-to-toe view of the model inside the environment from <Picture 4>. The model naturally enters the action or shifts into a confident presentation pose. Product A and Product B are already correctly worn, carried, attached, held, or positioned according to their real functions. The model briefly looks toward the camera and establishes a natural connection with the viewer.\n\n2.0–4.2 seconds — First side of the figure-eight:\nThe camera begins a smooth, wide figure-eight orbit around the complete model, moving toward one side while maintaining constant head-to-toe framing. The model takes a few natural steps, turns slightly, and performs an action that clearly demonstrates Product B. If it is wearable, the model gently turns to show its fit and movement. If it is carried or handheld, the model naturally lifts, adjusts, opens, or uses it while keeping it within the full-body composition.\n\n4.2–5.3 seconds — Figure-eight crossover:\nThe camera passes smoothly through the central crossover point in front of the model. The model performs a coordinated transition gesture involving both products, such as changing stance, transferring an item between hands, adjusting a wearable product, or turning the body. The complete person remains visible, and both products maintain correct size, identity, placement, and physical contact.\n\n5.3–7.8 seconds — Second side of the figure-eight:\nThe camera continues around the opposite side of the model, completing the second loop of the horizontal figure-eight path. Maintain the same wide full-body distance. The model performs a natural environment-appropriate interaction and demonstrates Product A through movement, orientation, or use. Show the products from a different angle without cutting to a product-only shot.\n\n7.8–10.0 seconds — Full-body hero finish:\nThe camera completes the figure-eight orbit and settles into a strong full-body hero composition. The model turns toward the camera and finishes in a confident, natural pose that clearly displays both Product A and Product B together. Keep the entire person, including both feet, fully visible. End with stable framing, a clear silhouette, and a polished commercial presentation.\n\nFIGURE-EIGHT CAMERA MOVEMENT\n\nThe camera follows one continuous, smooth horizontal figure-eight orbit centered on the complete model, not on either individual product.\n\nThe camera first curves around one side of the model, crosses gently through the front-center axis, then curves around the opposite side before returning to the hero angle.\n\nUse stabilized cinematic movement with gentle parallax, natural perspective, consistent camera height, moderate orbit radius, and subtle depth variation. The movement must feel like a controlled professional gimbal or camera crane.\n\nMaintain a wide full-body framing during the entire orbit. The camera may slightly adjust its height and distance to reveal product surfaces, but it must never move closer than a complete head-to-toe composition.\n\nDo not use product-only shots, macro shots, face-only shots, bust shots, waist-up shots, knee-up shots, extreme close-ups, abrupt zooms, rapid spins, sudden cuts, camera shaking, or disorienting motion.\n\nMODEL AND PRODUCT INTERACTION\n\nThe model behaves like a professional commercial presenter while remaining natural and physically believable.\n\nUse subtle eye contact, relaxed facial expressions, realistic breathing, natural blinking, gentle hair movement, accurate walking rhythm, balanced posture, anatomically correct hands, and realistic garment and object physics.\n\nThe model should interact directly with Product A and Product B through suitable actions such as wearing, carrying, holding, touching, adjusting, opening, turning, displaying, or using them.\n\nAll hand-to-product contact must be physically accurate. Fingers must wrap naturally around handles, straps, edges, controls, or surfaces. Products must react to gravity, body movement, fabric motion, momentum, and environmental airflow.\n\nIf additional people are present in the environment, they may perform subtle natural background actions and briefly react to the model. They must remain secondary and must not block, replace, or distract from the complete main model.\n\nLIGHT PARTICLE EFFECTS\n\nIntroduce refined, subtle light particles around the model and the two products. The particles gently follow the model’s movement, hand gestures, and the figure-eight camera flow.\n\nWhen the model touches, adjusts, carries, or uses a product, a small concentration of particles may briefly gather around the contact point and then disperse naturally.\n\nKeep the particle effects elegant, sparse, semi-transparent, and physically integrated into the scene. They should enhance the products without covering the model’s face, body, hands, feet, product structure, or environmental details.\n\nThe particles inherit their appearance naturally from the existing scene illumination and do not introduce a separate artificial color style.\n\nSCENE, ATMOSPHERE, AND VISUAL CONTINUITY\n\nPreserve the spatial scope, environmental atmosphere, lighting character, visual tone, color relationships, contrast, exposure, depth, material appearance, and overall mood established by <Picture 4> throughout the entire target video.\n\nMaintain the existing visual balance without assigning a new color style, introducing a separate color grade, or replacing the original lighting conditions.\n\nEnsure realistic reflections, contact shadows, depth, atmospheric perspective, environmental illumination, and consistent interaction between the model, the products, and the location.\n\nThe model and products must look physically present inside the environment rather than composited over it.\n\nIDENTITY AND OBJECT CONSISTENCY\n\nMaintain one consistent model identity throughout the entire video.\n\nMaintain one consistent instance of Product A and one consistent instance of Product B. Preserve their designs and prevent them from changing shape, color, size, material, orientation, or identity between frames.\n\nKeep all wearable products correctly fitted to the same body areas. Keep carryable and handheld products securely connected to the correct hands, straps, shoulders, or attachment points.\n\nPreserve realistic human anatomy, stable facial features, correct limb count, natural hand structure, accurate finger count, consistent body proportions, and physically plausible motion.\n\nSTRICT GENERATION REQUIREMENTS\n\nThe complete main model must remain visible from head to toe throughout the entire video.\nAlways show both feet with visible space beneath them.\nNever crop the head, hair, arms, hands, hips, legs, knees, ankles, or feet.\nNever replace the full-body composition with an isolated product shot.\nNever allow Product A or Product B to become the only subject in the frame.\nNever let products float, duplicate, disappear, merge with the body, deform, or change design.\nNever duplicate the model or create additional versions of the main model.\nNever import the product-image backgrounds into the final scene.\nNever replace the environment from <Picture 4> with a studio, runway, or unrelated location.\nNever add a new artificial color treatment or unrelated lighting style.\nNever allow particles to obscure the model or products.\nDo not generate new captions, subtitles, labels, interface elements, borders, watermarks, or unrelated text.\nDo not use abrupt transitions or separate catalogue-style insert shots.\n\nThe final result is a coherent, photorealistic, high-quality 10-second full-body commercial demonstration in which the model from <Picture 1> naturally presents and uses the automatically recognized products from <Picture 2> and <Picture 3> inside the environment from <Picture 4>, while a smooth figure-eight camera orbit and subtle interactive light particles enhance the presentation without sacrificing complete head-to-toe visibility."],"color":"#322","bgcolor":"#533"}],"links":[[250,120,0,121,1,"VAE"],[251,119,0,122,1,"VAE"],[253,129,0,125,0,"NOISE"],[254,126,0,125,1,"GUIDER"],[255,123,0,125,2,"SAMPLER"],[256,124,0,125,3,"SIGMAS"],[259,121,0,130,1,"AUDIO"],[260,130,0,92,0,"VIDEO"],[261,132,0,131,0,"FLOAT"],[270,136,0,126,1,"CONDITIONING"],[271,136,1,125,4,"LATENT"],[272,128,0,136,0,"CLIP"],[273,119,0,136,1,"VAE"],[274,120,0,136,2,"VAE"],[275,131,1,136,20,"INT"],[276,115,0,136,18,"INT"],[277,115,1,136,19,"INT"],[278,137,0,136,3,"IMAGE"],[279,138,0,136,17,"STRING"],[282,139,0,136,4,"IMAGE"],[283,141,0,136,5,"IMAGE"],[284,144,0,136,15,"AUDIO"],[285,122,0,147,0,"IMAGE"],[286,147,0,130,0,"IMAGE"],[287,149,0,148,0,"VIDEO"],[288,148,0,136,11,"IMAGE"],[289,148,1,136,13,"AUDIO"],[290,150,0,151,0,"MODEL"],[291,127,0,150,0,"MODEL"],[292,151,0,126,0,"MODEL"],[293,151,0,124,0,"MODEL"],[294,152,0,136,6,"IMAGE"],[295,153,0,136,7,"IMAGE"],[296,154,0,136,8,"IMAGE"],[297,155,0,136,9,"IMAGE"],[298,125,0,156,0,"LATENT"],[299,156,0,121,0,"LATENT"],[300,156,0,122,0,"LATENT"]],"groups":[{"id":1,"title":"Models","bounding":[-1520,4840,700,670],"color":"#3f789e","flags":{}},{"id":2,"title":"Sampling","bounding":[-280,4840,690,670],"color":"#3f789e","flags":{}},{"id":3,"title":"Conditioning","bounding":[-790,4840,480,670],"color":"#3f789e","flags":{}},{"id":4,"title":"Decoding and create video","bounding":[440,4840,670,670],"color":"#3f789e","flags":{}},{"id":5,"title":"User Inputs","bounding":[-1520,5540,1930,900],"color":"#3f789e","flags":{}},{"id":6,"title":"Group","bounding":[-994.8000489744444,4455.8261390551925,917.2272073278322,330],"color":"#3f789e","flags":{}}],"config":{},"extra":{"ds":{"scale":1.291467969000015,"offset":[2525.0177773559476,-5549.037373301024]},"frontendVersion":"1.47.11","VHS_latentpreview":false,"VHS_latentpreviewrate":0,"VHS_MetadataImage":true,"VHS_KeepIntermediate":true},"version":0.4}