ComfyPilotComfyPilot
工作流 / 视频 / VEED Lipsync 2.0：视频对口型
VEED Lipsync 2.0：视频对口型

VEED Lipsync 2.0：视频对口型

VEED Lipsync 2.0: Lip Sync

用 VEED Lipsync 2.0 把已有素材里说话人的口型替换成一段新的语音：画面中的人对着新音频说话，其余画面保持原样；成片保持原视频分辨率、25 fps，长度与音频一致，最长 600 秒。若视频比语音短，画面会正放再倒放直到音频结束。适合给现成镜头重新配音、不重拍就换台词，或制作本地化版本。

视频 5 个节点 2 个输入 81 次使用 ComfyUI ≥ 0.43.0 ComfyUI 6 KB

下载工作流 JSON 查看官方来源

下载后把文件拖进 ComfyUI 窗口即可打开。

节点构成

VeedLipsyncNode × 1
LoadVideo × 1
LoadAudio × 1
SaveVideo × 1
MarkdownNote × 1

官方标注的模型

VEED

工作流 JSON

和上面「下载工作流 JSON」是同一份文件。

1{
2 "id": "5b4ffc84-79bf-4ead-840c-689638e237ab",
3 "revision": 0,
4 "last_node_id": 11,
5 "last_link_id": 6,
6 "nodes": [
7 {
8 "id": 7,
9 "type": "VeedLipsyncNode",
10 "pos": [
11 2140,
12 730
13 ],
14 "size": [
15 320,
16 178
17 ],
18 "flags": [],
19 "order": 3,
20 "mode": 0,
21 "inputs": [
22 {
23 "name": "model.video",
24 "type": "VIDEO",
25 "link": 4
26 },
27 {
28 "name": "model.audio",
29 "type": "AUDIO",
30 "link": 5
31 }
32 ],
33 "outputs": [
34 {
35 "name": "VIDEO",
36 "type": "VIDEO",
37 "links": [
38 6
39 ]
40 }
41 ],
42 "properties": {
43 "Node name for S&R": "VeedLipsyncNode"
44 },
45 "widgets_values": [
46 "lipsync-2.0",
47 1576348213,
48 "randomize"
49 ],
50 "widgets_values_named": {
51 "model": "lipsync-2.0",
52 "model.seed": 1576348213,
53 "control_after_generate": "randomize"
54 },
55 "color": "#432",
56 "bgcolor": "#653"
57 },
58 {
59 "id": 8,
60 "type": "LoadVideo",
61 "pos": [
62 1610,
63 730
64 ],
65 "size": [
66 370,
67 460
68 ],
69 "flags": [],
70 "order": 0,
71 "mode": 0,
72 "inputs": [],
73 "outputs": [
74 {
75 "name": "VIDEO",
76 "type": "VIDEO",
77 "links": [
78 4
79 ]
80 }
81 ],
82 "properties": {
83 "Node name for S&R": "LoadVideo"
84 },
85 "widgets_values": [
86 "rainbow_face_portrait_prism_video.mp4",
87 "image"
88 ],
89 "widgets_values_named": {
90 "file": "rainbow_face_portrait_prism_video.mp4",
91 "upload": "image"
92 }
93 },
94 {
95 "id": 9,
96 "type": "LoadAudio",
97 "pos": [
98 1620,
99 1260
100 ],
101 "size": [
102 360,
103 162
104 ],
105 "flags": [],
106 "order": 1,
107 "mode": 0,
108 "inputs": [],
109 "outputs": [
110 {
111 "name": "AUDIO",
112 "type": "AUDIO",
113 "links": [
114 5
115 ]
116 }
117 ],
118 "properties": {
119 "Node name for S&R": "LoadAudio"
120 },
121 "widgets_values": [
122 "veed_lipsync_rainbow_night_speech.wav",
123 ""
124 ],
125 "widgets_values_named": {
126 "audio": "veed_lipsync_rainbow_night_speech.wav",
127 "upload": ""
128 }
129 },
130 {
131 "id": 10,
132 "type": "SaveVideo",
133 "pos": [
134 2520,
135 730
136 ],
137 "size": [
138 520,
139 160
140 ],
141 "flags": [],
142 "order": 4,
143 "mode": 0,
144 "inputs": [
145 {
146 "name": "video",
147 "type": "VIDEO",
148 "link": 6
149 }
150 ],
151 "outputs": [
152 {
153 "name": "video",
154 "type": "VIDEO",
155 "links": null
156 }
157 ],
158 "properties": [],
159 "widgets_values": [
160 "video/VEED_Lipsync_2.0",
161 "auto",
162 "auto",
163 "auto"
164 ],
165 "widgets_values_named": {
166 "filename_prefix": "video/VEED_Lipsync_2.0",
167 "format": "auto",
168 "format.codec": "auto",
169 "codec": "auto"
170 }
171 },
172 {
173 "id": 11,
174 "type": "MarkdownNote",
175 "pos": [
176 1040,
177 730
178 ],
179 "size": [
180 520,
181 620
182 ],
183 "flags": [],
184 "order": 2,
185 "mode": 0,
186 "inputs": [],
187 "outputs": [],
188 "properties": [],
189 "widgets_values": [
190 "## VEED Lipsync 2.0: Video Lip Sync — Workflow Guide\n\n## Input Assets\n- [rainbow_face_portrait_prism_video.mp4](https://raw.githubusercontent.com/Comfy-Org/workflow_templates/refs/heads/main/input/rainbow_face_portrait_prism_video.mp4)\n- [veed_lipsync_rainbow_night_speech.wav](https://raw.githubusercontent.com/Comfy-Org/workflow_templates/refs/heads/main/input/veed_lipsync_rainbow_night_speech.wav)\n\n## How to Use\n1. Load footage of one front-facing speaker.\n2. Load the new speech; it sets the clip length.\n3. Run; Save Video writes the result.\n\n## Limits\n- Audio up to 600 seconds; output length matches it.\n- Output keeps the source resolution at 25 fps.\n- Shorter video than audio plays forward then backward until it ends.\n"
191 ],
192 "widgets_values_named": {
193 "text": "## VEED Lipsync 2.0: Video Lip Sync — Workflow Guide\n\n## Input Assets\n- [rainbow_face_portrait_prism_video.mp4](https://raw.githubusercontent.com/Comfy-Org/workflow_templates/refs/heads/main/input/rainbow_face_portrait_prism_video.mp4)\n- [veed_lipsync_rainbow_night_speech.wav](https://raw.githubusercontent.com/Comfy-Org/workflow_templates/refs/heads/main/input/veed_lipsync_rainbow_night_speech.wav)\n\n## How to Use\n1. Load footage of one front-facing speaker.\n2. Load the new speech; it sets the clip length.\n3. Run; Save Video writes the result.\n\n## Limits\n- Audio up to 600 seconds; output length matches it.\n- Output keeps the source resolution at 25 fps.\n- Shorter video than audio plays forward then backward until it ends.\n"
194 },
195 "color": "#222",
196 "bgcolor": "#000",
197 "title": "Note: Workflow Guide"
198 }
199 ],
200 "links": [
201 [
202 4,
203 8,
204 0,
205 7,
206 0,
207 "VIDEO"
208 ],
209 [
210 5,
211 9,
212 0,
213 7,
214 1,
215 "AUDIO"
216 ],
217 [
218 6,
219 7,
220 0,
221 10,
222 0,
223 "VIDEO"
224 ]
225 ],
226 "groups": [],
227 "config": [],
228 "extra": {
229 "ds": {
230 "scale": 0.9377531531912865,
231 "offset": [
232 -891.9774219531085,
233 -335.56155794770086
234 ]
235 },
236 "frontendVersion": "1.58.0"
237 },
238 "version": 0.4
239}