工作流 /
视频 /
混元 视频 图像 转 视频（混元 视频）
混元 视频 图像 转 视频（混元 视频）
hunyuan video image to video
ComfyUI 官方示例：混元 视频 功能演示，可拖进 ComfyUI 直接使用。
视频
24 个节点 Comfy-Org 20 KB
需要的权重
这个工作流引用了 5 个权重文件，其中 5 个能在本站模型库里找到。
clip_l.safetensors
在模型库查看
hunyuan_video_image_to_video_720p_bf16.safetensors
在模型库查看
hunyuan_video_vae_bf16.safetensors
在模型库查看
llava_llama3_fp8_scaled.safetensors
在模型库查看
llava_llama3_vision.safetensors
在模型库查看
节点构成
Note
× 4
BasicGuider
× 1
VAEDecode
× 1
SamplerCustomAdvanced
× 1
FluxGuidance
× 1
ModelSamplingSD3
× 1
BasicScheduler
× 1
KSamplerSelect
× 1
HunyuanImageToVideo
× 1
VAEDecodeTiled
× 1
SaveAnimatedWEBP
× 1
SaveWEBM
× 1
TextEncodeHunyuanVideo_ImageToVideo
× 1
CLIPVisionEncode
× 1
LoadImage
× 1
CLIPVisionLoader
× 1
官方标注的模型
clip_l.safetensorshunyuan_video_image_to_video_720p_bf16.safetensorshunyuan_video_vae_bf16.safetensorsllava_llama3_fp8_scaled.safetensorsllava_llama3_vision.safetensors
工作流 JSON
和上面「下载工作流 JSON」是同一份文件。
1{
2 "last_node_id": 89,
3 "last_link_id": 230,
4 "nodes": [
5 {
6 "id": 22,
7 "type": "BasicGuider",
8 "pos": [
9 600,
10 0
11 ],
12 "size": [
13 222.3482666015625,
14 46
15 ],
16 "flags": [],
17 "order": 17,
18 "mode": 0,
19 "inputs": [
20 {
21 "name": "model",
22 "type": "MODEL",
23 "link": 195,
24 "slot_index": 0
25 },
26 {
27 "name": "conditioning",
28 "type": "CONDITIONING",
29 "link": 129,
30 "slot_index": 1
31 }
32 ],
33 "outputs": [
34 {
35 "name": "GUIDER",
36 "type": "GUIDER",
37 "shape": 3,
38 "links": [
39 30
40 ],
41 "slot_index": 0
42 }
43 ],
44 "properties": {
45 "Node name for S&R": "BasicGuider"
46 },
47 "widgets_values": []
48 },
49 {
50 "id": 8,
51 "type": "VAEDecode",
52 "pos": [
53 1150,
54 90
55 ],
56 "size": [
57 210,
58 46
59 ],
60 "flags": [],
61 "order": 19,
62 "mode": 2,
63 "inputs": [
64 {
65 "name": "samples",
66 "type": "LATENT",
67 "link": 181
68 },
69 {
70 "name": "vae",
71 "type": "VAE",
72 "link": 206
73 }
74 ],
75 "outputs": [
76 {
77 "name": "IMAGE",
78 "type": "IMAGE",
79 "links": [],
80 "slot_index": 0
81 }
82 ],
83 "properties": {
84 "Node name for S&R": "VAEDecode"
85 },
86 "widgets_values": []
87 },
88 {
89 "id": 13,
90 "type": "SamplerCustomAdvanced",
91 "pos": [
92 860,
93 200
94 ],
95 "size": [
96 272.3617858886719,
97 124.53733825683594
98 ],
99 "flags": [],
100 "order": 18,
101 "mode": 0,
102 "inputs": [
103 {
104 "name": "noise",
105 "type": "NOISE",
106 "link": 37,
107 "slot_index": 0
108 },
109 {
110 "name": "guider",
111 "type": "GUIDER",
112 "link": 30,
113 "slot_index": 1
114 },
115 {
116 "name": "sampler",
117 "type": "SAMPLER",
118 "link": 19,
119 "slot_index": 2
120 },
121 {
122 "name": "sigmas",
123 "type": "SIGMAS",
124 "link": 20,
125 "slot_index": 3
126 },
127 {
128 "name": "latent_image",
129 "type": "LATENT",
130 "link": 216,
131 "slot_index": 4
132 }
133 ],
134 "outputs": [
135 {
136 "name": "output",
137 "type": "LATENT",
138 "shape": 3,
139 "links": [
140 181,
141 210
142 ],
143 "slot_index": 0
144 },
145 {
146 "name": "denoised_output",
147 "type": "LATENT",
148 "shape": 3,
149 "links": null
150 }
151 ],
152 "properties": {
153 "Node name for S&R": "SamplerCustomAdvanced"
154 },
155 "widgets_values": []
156 },
157 {
158 "id": 74,
159 "type": "Note",
160 "pos": [
161 1147.7459716796875,
162 405.0789489746094
163 ],
164 "size": [
165 210,
166 170
167 ],
168 "flags": [],
169 "order": 0,
170 "mode": 0,
171 "inputs": [],
172 "outputs": [],
173 "properties": [],
174 "widgets_values": [
175 "Use the tiled decode node by default because most people will need it.\n\nLower the tile_size and overlap if you run out of memory."
176 ],
177 "color": "#432",
178 "bgcolor": "#653"
179 },
180 {
181 "id": 26,
182 "type": "FluxGuidance",
183 "pos": [
184 514.2149047851562,
185 86.77685546875
186 ],
187 "size": [
188 317.4000244140625,
189 58
190 ],
191 "flags": [],
192 "order": 16,
193 "mode": 0,
194 "inputs": [
195 {
196 "name": "conditioning",
197 "type": "CONDITIONING",
198 "link": 225
199 }
200 ],
201 "outputs": [
202 {
203 "name": "CONDITIONING",
204 "type": "CONDITIONING",
205 "shape": 3,
206 "links": [
207 129
208 ],
209 "slot_index": 0
210 }
211 ],
212 "properties": {
213 "Node name for S&R": "FluxGuidance"
214 },
215 "widgets_values": [
216 6
217 ],
218 "color": "#233",
219 "bgcolor": "#355"
220 },
221 {
222 "id": 67,
223 "type": "ModelSamplingSD3",
224 "pos": [
225 360,
226 0
227 ],
228 "size": [
229 210,
230 58
231 ],
232 "flags": [],
233 "order": 13,
234 "mode": 0,
235 "inputs": [
236 {
237 "name": "model",
238 "type": "MODEL",
239 "link": 209
240 }
241 ],
242 "outputs": [
243 {
244 "name": "MODEL",
245 "type": "MODEL",
246 "links": [
247 195
248 ],
249 "slot_index": 0
250 }
251 ],
252 "properties": {
253 "Node name for S&R": "ModelSamplingSD3"
254 },
255 "widgets_values": [
256 7
257 ]
258 },
259 {
260 "id": 17,
261 "type": "BasicScheduler",
262 "pos": [
263 510,
264 660
265 ],
266 "size": [
267 315,
268 106
269 ],
270 "flags": [],
271 "order": 12,
272 "mode": 0,
273 "inputs": [
274 {
275 "name": "model",
276 "type": "MODEL",
277 "link": 190,
278 "slot_index": 0
279 }
280 ],
281 "outputs": [
282 {
283 "name": "SIGMAS",
284 "type": "SIGMAS",
285 "shape": 3,
286 "links": [
287 20
288 ]
289 }
290 ],
291 "properties": {
292 "Node name for S&R": "BasicScheduler"
293 },
294 "widgets_values": [
295 "simple",
296 20,
297 1
298 ]
299 },
300 {
301 "id": 16,
302 "type": "KSamplerSelect",
303 "pos": [
304 520,
305 550
306 ],
307 "size": [
308 315,
309 58
310 ],
311 "flags": [],
312 "order": 1,
313 "mode": 0,
314 "inputs": [],
315 "outputs": [
316 {
317 "name": "SAMPLER",
318 "type": "SAMPLER",
319 "shape": 3,
320 "links": [
321 19
322 ]
323 }
324 ],
325 "properties": {
326 "Node name for S&R": "KSamplerSelect"
327 },
328 "widgets_values": [
329 "euler"
330 ]
331 },
332 {
333 "id": 78,
334 "type": "HunyuanImageToVideo",
335 "pos": [
336 510,
337 820
338 ],
339 "size": [
340 315,
341 170
342 ],
343 "flags": [],
344 "order": 15,
345 "mode": 0,
346 "inputs": [
347 {
348 "name": "positive",
349 "type": "CONDITIONING",
350 "link": 218
351 },
352 {
353 "name": "vae",
354 "type": "VAE",
355 "link": 223
356 },
357 {
358 "name": "start_image",
359 "type": "IMAGE",
360 "shape": 7,
361 "link": 222
362 }
363 ],
364 "outputs": [
365 {
366 "name": "positive",
367 "type": "CONDITIONING",
368 "links": [
369 225
370 ],
371 "slot_index": 0
372 },
373 {
374 "name": "latent",
375 "type": "LATENT",
376 "links": [
377 216
378 ]
379 }
380 ],
381 "properties": {
382 "Node name for S&R": "HunyuanImageToVideo"
383 },
384 "widgets_values": [
385 720,
386 720,
387 57,
388 1
389 ]
390 },
391 {
392 "id": 73,
393 "type": "VAEDecodeTiled",
394 "pos": [
395 1150,
396 200
397 ],
398 "size": [
399 210,
400 150
401 ],
402 "flags": [],
403 "order": 20,
404 "mode": 0,
405 "inputs": [
406 {
407 "name": "samples",
408 "type": "LATENT",
409 "link": 210
410 },
411 {
412 "name": "vae",
413 "type": "VAE",
414 "link": 211
415 }
416 ],
417 "outputs": [
418 {
419 "name": "IMAGE",
420 "type": "IMAGE",
421 "links": [
422 230
423 ],
424 "slot_index": 0
425 }
426 ],
427 "properties": {
428 "Node name for S&R": "VAEDecodeTiled"
429 },
430 "widgets_values": [
431 256,
432 64,
433 64,
434 8
435 ]
436 },
437 {
438 "id": 75,
439 "type": "SaveAnimatedWEBP",
440 "pos": [
441 1640,
442 200
443 ],
444 "size": [
445 621.495361328125,
446 587.12451171875
447 ],
448 "flags": [],
449 "order": 22,
450 "mode": 0,
451 "inputs": [
452 {
453 "name": "images",
454 "type": "IMAGE",
455 "link": 228
456 }
457 ],
458 "outputs": [],
459 "properties": [],
460 "widgets_values": [
461 "ComfyUI",
462 24,
463 false,
464 90,
465 "default"
466 ]
467 },
468 {
469 "id": 84,
470 "type": "SaveWEBM",
471 "pos": [
472 2280,
473 200
474 ],
475 "size": [
476 315,
477 130
478 ],
479 "flags": [],
480 "order": 23,
481 "mode": 4,
482 "inputs": [
483 {
484 "name": "images",
485 "type": "IMAGE",
486 "link": 229
487 }
488 ],
489 "outputs": [],
490 "properties": {
491 "Node name for S&R": "SaveWEBM"
492 },
493 "widgets_values": [
494 "ComfyUI",
495 "vp9",
496 24,
497 12
498 ]
499 },
500 {
501 "id": 87,
502 "type": "Note",
503 "pos": [
504 1410,
505 340
506 ],
507 "size": [
508 210,
509 170
510 ],
511 "flags": [],
512 "order": 2,
513 "mode": 0,
514 "inputs": [],
515 "outputs": [],
516 "properties": [],
517 "widgets_values": [
518 "The official code removes the first 4 frames because they are sometimes bad. You can bypass (CTRL-B) this node if you don't want this."
519 ],
520 "color": "#432",
521 "bgcolor": "#653"
522 },
523 {
524 "id": 80,
525 "type": "TextEncodeHunyuanVideo_ImageToVideo",
526 "pos": [
527 390,
528 180
529 ],
530 "size": [
531 441,
532 200
533 ],
534 "flags": [],
535 "order": 14,
536 "mode": 0,
537 "inputs": [
538 {
539 "name": "clip",
540 "type": "CLIP",
541 "link": 224
542 },
543 {
544 "name": "clip_vision_output",
545 "type": "CLIP_VISION_OUTPUT",
546 "link": 219
547 }
548 ],
549 "outputs": [
550 {
551 "name": "CONDITIONING",
552 "type": "CONDITIONING",
553 "links": [
554 218
555 ]
556 }
557 ],
558 "title": "Text Encode Hunyuan Video (ImageToVideo)",
559 "properties": {
560 "Node name for S&R": "TextEncodeHunyuanVideo_ImageToVideo"
561 },
562 "widgets_values": [
563 "a cute anime girl with massive fennec ears and a big fluffy tail wearing a maid outfit running"
564 ],
565 "color": "#232",
566 "bgcolor": "#353"
567 },
568 {
569 "id": 81,
570 "type": "CLIPVisionEncode",
571 "pos": [
572 200,
573 530
574 ],
575 "size": [
576 253.60000610351562,
577 78
578 ],
579 "flags": [],
580 "order": 11,
581 "mode": 0,
582 "inputs": [
583 {
584 "name": "clip_vision",
585 "type": "CLIP_VISION",
586 "link": 220
587 },
588 {
589 "name": "image",
590 "type": "IMAGE",
591 "link": 221
592 }
593 ],
594 "outputs": [
595 {
596 "name": "CLIP_VISION_OUTPUT",
597 "type": "CLIP_VISION_OUTPUT",
598 "links": [
599 219
600 ]
601 }
602 ],
603 "properties": {
604 "Node name for S&R": "CLIPVisionEncode"
605 },
606 "widgets_values": [
607 "none"
608 ]
609 },
610 {
611 "id": 89,
612 "type": "Note",
613 "pos": [
614 190,
615 660
616 ],
617 "size": [
618 260,
619 210
620 ],
621 "flags": [],
622 "order": 3,
623 "mode": 0,
624 "inputs": [],
625 "outputs": [],
626 "properties": [],
627 "widgets_values": [
628 "Select your image here. The image is fed to both the text encoder and directly to the model.\n\nYou can set the resolution and length of the video using the HunyuanImageToVideo node."
629 ],
630 "color": "#432",
631 "bgcolor": "#653"
632 },
633 {
634 "id": 83,
635 "type": "LoadImage",
636 "pos": [
637 -190,
638 700
639 ],
640 "size": [
641 365.4132080078125,
642 471.8512268066406
643 ],
644 "flags": [],
645 "order": 4,
646 "mode": 0,
647 "inputs": [],
648 "outputs": [
649 {
650 "name": "IMAGE",
651 "type": "IMAGE",
652 "links": [
653 221,
654 222
655 ],
656 "slot_index": 0
657 },
658 {
659 "name": "MASK",
660 "type": "MASK",
661 "links": null
662 }
663 ],
664 "properties": {
665 "Node name for S&R": "LoadImage"
666 },
667 "widgets_values": [
668 "flux_dev_example.png",
669 "image"
670 ]
671 },
672 {
673 "id": 82,
674 "type": "CLIPVisionLoader",
675 "pos": [
676 -190,
677 580
678 ],
679 "size": [
680 315,
681 58
682 ],
683 "flags": [],
684 "order": 5,
685 "mode": 0,
686 "inputs": [],
687 "outputs": [
688 {
689 "name": "CLIP_VISION",
690 "type": "CLIP_VISION",
691 "links": [
692 220
693 ]
694 }
695 ],
696 "properties": {
697 "Node name for S&R": "CLIPVisionLoader"
698 },
699 "widgets_values": [
700 "llava_llama3_vision.safetensors"
701 ],
702 "color": "#223",
703 "bgcolor": "#335"
704 },
705 {
706 "id": 10,
707 "type": "VAELoader",
708 "pos": [
709 -190,
710 470
711 ],
712 "size": [
713 350,
714 60
715 ],
716 "flags": [],
717 "order": 6,
718 "mode": 0,
719 "inputs": [],
720 "outputs": [
721 {
722 "name": "VAE",
723 "type": "VAE",
724 "shape": 3,
725 "links": [
726 206,
727 211,
728 223
729 ],
730 "slot_index": 0
731 }
732 ],
733 "properties": {
734 "Node name for S&R": "VAELoader"
735 },
736 "widgets_values": [
737 "hunyuan_video_vae_bf16.safetensors"
738 ],
739 "color": "#223",
740 "bgcolor": "#335"
741 },
742 {
743 "id": 11,
744 "type": "DualCLIPLoader",
745 "pos": [
746 -190,
747 290
748 ],
749 "size": [
750 350,
751 122
752 ],
753 "flags": [],
754 "order": 7,
755 "mode": 0,
756 "inputs": [],
757 "outputs": [
758 {
759 "name": "CLIP",
760 "type": "CLIP",
761 "shape": 3,
762 "links": [
763 224
764 ],
765 "slot_index": 0
766 }
767 ],
768 "properties": {
769 "Node name for S&R": "DualCLIPLoader"
770 },
771 "widgets_values": [
772 "clip_l.safetensors",
773 "llava_llama3_fp8_scaled.safetensors",
774 "hunyuan_video",
775 "default"
776 ],
777 "color": "#223",
778 "bgcolor": "#335"
779 },
780 {
781 "id": 12,
782 "type": "UNETLoader",
783 "pos": [
784 -190,
785 160
786 ],
787 "size": [
788 404.6181640625,
789 82
790 ],
791 "flags": [],
792 "order": 8,
793 "mode": 0,
794 "inputs": [],
795 "outputs": [
796 {
797 "name": "MODEL",
798 "type": "MODEL",
799 "shape": 3,
800 "links": [
801 190,
802 209
803 ],
804 "slot_index": 0
805 }
806 ],
807 "properties": {
808 "Node name for S&R": "UNETLoader"
809 },
810 "widgets_values": [
811 "hunyuan_video_image_to_video_720p_bf16.safetensors",
812 "default"
813 ],
814 "color": "#223",
815 "bgcolor": "#335"
816 },
817 {
818 "id": 77,
819 "type": "Note",
820 "pos": [
821 -140,
822 0
823 ],
824 "size": [
825 350,
826 110
827 ],
828 "flags": [],
829 "order": 9,
830 "mode": 0,
831 "inputs": [],
832 "outputs": [],
833 "properties": [],
834 "widgets_values": [
835 "Select a fp8 weight_dtype if you are running out of memory."
836 ],
837 "color": "#432",
838 "bgcolor": "#653"
839 },
840 {
841 "id": 86,
842 "type": "ImageFromBatch",
843 "pos": [
844 1410,
845 200
846 ],
847 "size": [
848 210,
849 82
850 ],
851 "flags": [],
852 "order": 21,
853 "mode": 0,
854 "inputs": [
855 {
856 "name": "image",
857 "type": "IMAGE",
858 "link": 230
859 }
860 ],
861 "outputs": [
862 {
863 "name": "IMAGE",
864 "type": "IMAGE",
865 "links": [
866 228,
867 229
868 ]
869 }
870 ],
871 "properties": {
872 "Node name for S&R": "ImageFromBatch"
873 },
874 "widgets_values": [
875 4,
876 4096
877 ]
878 },
879 {
880 "id": 25,
881 "type": "RandomNoise",
882 "pos": [
883 520,
884 420
885 ],
886 "size": [
887 315,
888 82
889 ],
890 "flags": [],
891 "order": 10,
892 "mode": 0,
893 "inputs": [],
894 "outputs": [
895 {
896 "name": "NOISE",
897 "type": "NOISE",
898 "shape": 3,
899 "links": [
900 37
901 ]
902 }
903 ],
904 "properties": {
905 "Node name for S&R": "RandomNoise"
906 },
907 "widgets_values": [
908 187330947843879,
909 "randomize"
910 ],
911 "color": "#2a363b",
912 "bgcolor": "#3f5159"
913 }
914 ],
915 "links": [
916 [
917 19,
918 16,
919 0,
920 13,
921 2,
922 "SAMPLER"
923 ],
924 [
925 20,
926 17,
927 0,
928 13,
929 3,
930 "SIGMAS"
931 ],
932 [
933 30,
934 22,
935 0,
936 13,
937 1,
938 "GUIDER"
939 ],
940 [
941 37,
942 25,
943 0,
944 13,
945 0,
946 "NOISE"
947 ],
948 [
949 129,
950 26,
951 0,
952 22,
953 1,
954 "CONDITIONING"
955 ],
956 [
957 181,
958 13,
959 0,
960 8,
961 0,
962 "LATENT"
963 ],
964 [
965 190,
966 12,
967 0,
968 17,
969 0,
970 "MODEL"
971 ],
972 [
973 195,
974 67,
975 0,
976 22,
977 0,
978 "MODEL"
979 ],
980 [
981 206,
982 10,
983 0,
984 8,
985 1,
986 "VAE"
987 ],
988 [
989 209,
990 12,
991 0,
992 67,
993 0,
994 "MODEL"
995 ],
996 [
997 210,
998 13,
999 0,
1000 73,
1001 0,
1002 "LATENT"
1003 ],
1004 [
1005 211,
1006 10,
1007 0,
1008 73,
1009 1,
1010 "VAE"
1011 ],
1012 [
1013 216,
1014 78,
1015 1,
1016 13,
1017 4,
1018 "LATENT"
1019 ],
1020 [
1021 218,
1022 80,
1023 0,
1024 78,
1025 0,
1026 "CONDITIONING"
1027 ],
1028 [
1029 219,
1030 81,
1031 0,
1032 80,
1033 1,
1034 "CLIP_VISION_OUTPUT"
1035 ],
1036 [
1037 220,
1038 82,
1039 0,
1040 81,
1041 0,
1042 "CLIP_VISION"
1043 ],
1044 [
1045 221,
1046 83,
1047 0,
1048 81,
1049 1,
1050 "IMAGE"
1051 ],
1052 [
1053 222,
1054 83,
1055 0,
1056 78,
1057 2,
1058 "IMAGE"
1059 ],
1060 [
1061 223,
1062 10,
1063 0,
1064 78,
1065 1,
1066 "VAE"
1067 ],
1068 [
1069 224,
1070 11,
1071 0,
1072 80,
1073 0,
1074 "CLIP"
1075 ],
1076 [
1077 225,
1078 78,
1079 0,
1080 26,
1081 0,
1082 "CONDITIONING"
1083 ],
1084 [
1085 228,
1086 86,
1087 0,
1088 75,
1089 0,
1090 "IMAGE"
1091 ],
1092 [
1093 229,
1094 86,
1095 0,
1096 84,
1097 0,
1098 "IMAGE"
1099 ],
1100 [
1101 230,
1102 73,
1103 0,
1104 86,
1105 0,
1106 "IMAGE"
1107 ]
1108 ],
1109 "groups": [],
1110 "config": [],
1111 "extra": {
1112 "groupNodes": []
1113 },
1114 "models": [
1115 {
1116 "name": "llava_llama3_vision.safetensors",
1117 "url": "https://huggingface.co/Comfy-Org/HunyuanVideo_repackaged/resolve/main/split_files/clip_vision/llava_llama3_vision.safetensors?download=true",
1118 "directory": "clip_vision"
1119 },
1120 {
1121 "name": "clip_l.safetensors",
1122 "url": "https://huggingface.co/Comfy-Org/HunyuanVideo_repackaged/resolve/main/split_files/text_encoders/clip_l.safetensors?download=true",
1123 "directory": "text_encoders"
1124 },
1125 {
1126 "name": "llava_llama3_fp8_scaled.safetensors",
1127 "url": "https://huggingface.co/Comfy-Org/HunyuanVideo_repackaged/resolve/main/split_files/text_encoders/llava_llama3_fp8_scaled.safetensors?download=true",
1128 "directory": "text_encoders"
1129 },
1130 {
1131 "name": "hunyuan_video_vae_bf16.safetensors",
1132 "url": "https://huggingface.co/Comfy-Org/HunyuanVideo_repackaged/resolve/main/split_files/vae/hunyuan_video_vae_bf16.safetensors?download=true",
1133 "directory": "vae"
1134 },
1135 {
1136 "name": "hunyuan_video_image_to_video_720p_bf16.safetensors",
1137 "url": "https://huggingface.co/Comfy-Org/HunyuanVideo_repackaged/resolve/main/split_files/diffusion_models/hunyuan_video_image_to_video_720p_bf16.safetensors?download=true",
1138 "directory": "diffusion_models"
1139 }
1140 ],
1141 "version": 0.4
1142}