工作流 /
音频 /
Chatter Box：使用TTS进行语音克隆
Chatter Box：使用TTS进行语音克隆
Chatter Box: Voice Cloning with TTS
上传一段简短语音片段并输入文本，以克隆的声音生成新的语音。
音频
4 个节点 1 个输入 73 次使用 ComfyUI ≥ 0.25.0 PurzBeats 6 KB
节点构成
FL_ChatterboxTTS
× 1
MarkdownNote
× 1
LoadAudio
× 1
SaveAudioAdvanced
× 1
官方标注的模型
Chatter Box
工作流 JSON
和上面「下载工作流 JSON」是同一份文件。
1{
2 "id": "f0d08102-91a0-4ba1-b166-0dd01f7e5aa9",
3 "revision": 0,
4 "last_node_id": 8,
5 "last_link_id": 8,
6 "nodes": [
7 {
8 "id": 4,
9 "type": "FL_ChatterboxTTS",
10 "pos": [
11 -170,
12 240
13 ],
14 "size": [
15 380,
16 320
17 ],
18 "flags": [],
19 "order": 2,
20 "mode": 0,
21 "inputs": [
22 {
23 "name": "audio_prompt",
24 "shape": 7,
25 "type": "AUDIO",
26 "link": 7
27 }
28 ],
29 "outputs": [
30 {
31 "name": "audio",
32 "type": "AUDIO",
33 "links": [
34 8
35 ]
36 },
37 {
38 "name": "message",
39 "type": "STRING",
40 "links": null
41 }
42 ],
43 "properties": {
44 "Node name for S&R": "FL_ChatterboxTTS"
45 },
46 "widgets_values": [
47 "This is Chatterbox—and it just recreated my voice from a nine-second recording.\n\nGive it a shot with your own voice and hear the results for yourself.",
48 0.5,
49 0.5,
50 0.8,
51 390519100,
52 "randomize",
53 false,
54 false
55 ],
56 "widgets_values_named": {
57 "text": "This is Chatterbox—and it just recreated my voice from a nine-second recording.\n\nGive it a shot with your own voice and hear the results for yourself.",
58 "exaggeration": 0.5,
59 "cfg_weight": 0.5,
60 "temperature": 0.8,
61 "seed": 390519100,
62 "control_after_generate": "randomize",
63 "use_cpu": false,
64 "keep_model_loaded": false
65 },
66 "color": "#16727c",
67 "bgcolor": "#4F0074"
68 },
69 {
70 "id": 7,
71 "type": "MarkdownNote",
72 "pos": [
73 -990,
74 240
75 ],
76 "size": [
77 470,
78 560
79 ],
80 "flags": [],
81 "order": 0,
82 "mode": 0,
83 "inputs": [],
84 "outputs": [],
85 "title": "🎙️ Chatterbox Text-To-Speech",
86 "properties": [],
87 "widgets_values": [
88 "Clone a voice from a short audio clip and generate new speech using a simple text prompt.\n\n---\n\n## Quick Start\n\n**Input:** short voice clip \n**Output:** generated speech in the same voice\n\n---\n\n## Steps\n\n### 1. Load a Voice\n- Add an audio file to the **Load Audio** node \n- Even a **5–10 second** clip is enough \n- This clip is used as the voice reference\n\n### 2. Enter Text\n- Type what you want spoken in the **FL Chatterbox TTS** node \n- Write it as you’d naturally say it—pauses and punctuation matter\n\n### 3. (Optional) Tweak Controls\n- **Exaggeration** — increases expressiveness \n- **CFG Weight** — how tightly the voice follows the prompt \n- **Temperature** — adds variation and naturalness \n_Default settings work well for first runs_\n\n### 4. Generate\n- Click **Run** \n- The generated audio will appear in **Save Audio (Advanced)** (FLAC)\n\n---\n\n## Notes\n- Clean, single-speaker recordings give the best results \n- Background noise can affect voice accuracy \n- Try small text changes to hear different inflections\n\n---\n"
89 ],
90 "widgets_values_named": {
91 "text": "Clone a voice from a short audio clip and generate new speech using a simple text prompt.\n\n---\n\n## Quick Start\n\n**Input:** short voice clip \n**Output:** generated speech in the same voice\n\n---\n\n## Steps\n\n### 1. Load a Voice\n- Add an audio file to the **Load Audio** node \n- Even a **5–10 second** clip is enough \n- This clip is used as the voice reference\n\n### 2. Enter Text\n- Type what you want spoken in the **FL Chatterbox TTS** node \n- Write it as you’d naturally say it—pauses and punctuation matter\n\n### 3. (Optional) Tweak Controls\n- **Exaggeration** — increases expressiveness \n- **CFG Weight** — how tightly the voice follows the prompt \n- **Temperature** — adds variation and naturalness \n_Default settings work well for first runs_\n\n### 4. Generate\n- Click **Run** \n- The generated audio will appear in **Save Audio (Advanced)** (FLAC)\n\n---\n\n## Notes\n- Clean, single-speaker recordings give the best results \n- Background noise can affect voice accuracy \n- Try small text changes to hear different inflections\n\n---\n"
92 },
93 "color": "#222",
94 "bgcolor": "#000"
95 },
96 {
97 "id": 6,
98 "type": "LoadAudio",
99 "pos": [
100 -480,
101 240
102 ],
103 "size": [
104 290,
105 140
106 ],
107 "flags": [],
108 "order": 1,
109 "mode": 0,
110 "inputs": [],
111 "outputs": [
112 {
113 "name": "AUDIO",
114 "type": "AUDIO",
115 "links": [
116 7
117 ]
118 }
119 ],
120 "properties": {
121 "Node name for S&R": "LoadAudio"
122 },
123 "widgets_values": [
124 "chatterbox_input_target_voice.mp3",
125 null
126 ],
127 "widgets_values_named": {
128 "audio": "chatterbox_input_target_voice.mp3",
129 "upload": null
130 }
131 },
132 {
133 "id": 8,
134 "type": "SaveAudioAdvanced",
135 "pos": [
136 240,
137 240
138 ],
139 "size": [
140 380,
141 149.99997158017334
142 ],
143 "flags": [],
144 "order": 3,
145 "mode": 0,
146 "inputs": [
147 {
148 "name": "audio",
149 "type": "AUDIO",
150 "link": 8
151 }
152 ],
153 "outputs": [
154 {
155 "name": "audio",
156 "type": "AUDIO",
157 "links": null
158 }
159 ],
160 "properties": [],
161 "widgets_values": [
162 "audio/chatterbox_tts",
163 "flac"
164 ],
165 "widgets_values_named": {
166 "filename_prefix": "audio/chatterbox_tts",
167 "format": "flac"
168 }
169 }
170 ],
171 "links": [
172 [
173 7,
174 6,
175 0,
176 4,
177 0,
178 "AUDIO"
179 ],
180 [
181 8,
182 4,
183 0,
184 8,
185 0,
186 "AUDIO"
187 ]
188 ],
189 "groups": [],
190 "config": [],
191 "extra": {
192 "ds": {
193 "scale": 0.5474789080190482,
194 "offset": [
195 1660.6001336655434,
196 239.71937655397608
197 ]
198 },
199 "frontendVersion": "1.54.15",
200 "workflowRendererVersion": "LG",
201 "VHS_latentpreview": false,
202 "VHS_latentpreviewrate": 0,
203 "VHS_MetadataImage": true,
204 "VHS_KeepIntermediate": true
205 },
206 "version": 0.4
207}