{ "config": {}, "extra": { "VHS_KeepIntermediate": true, "VHS_MetadataImage": true, "VHS_latentpreview": false, "VHS_latentpreviewrate": 0, "ds": { "offset": [ -708.619555879183, -356.28310857753746 ], "scale": 0.8376431567963556 }, "frontendVersion": "1.56.0", "links_added_by_ue": [], "ue_links": [], "workflowRendererVersion": "Vue-corrected" }, "groups": [ { "bounding": [ 1790, 870, 600, 510 ], "color": "#3f789e", "flags": {}, "id": 1, "title": "Option 2: Create a Voice Clone" }, { "bounding": [ 1790, 660, 600, 200 ], "color": "#3f789e", "flags": {}, "id": 2, "title": "Option 1: Use Preset Voice" } ], "id": "91f6bbe2-ed41-4fd6-bac7-71d5b5864ecb", "last_link_id": 300, "last_node_id": 219, "links": [ [ 300, 211, 0, 208, 0, "ELEVENLABS_VOICE" ], [ 296, 210, 0, 212, 0, "AUDIO" ], [ 298, 208, 0, 218, 0, "AUDIO" ] ], "nodes": [ { "bgcolor": "#653", "color": "#432", "flags": {}, "id": 208, "inputs": [ { "link": 300, "name": "voice", "type": "ELEVENLABS_VOICE" } ], "mode": 0, "order": 6, "outputs": [ { "links": [ 298 ], "name": "AUDIO", "type": "AUDIO" } ], "pos": [ 2430, 730 ], "properties": { "Node name for S&R": "ElevenLabsTextToSpeech", "cnr_id": "comfy-core", "ver": "0.14.1" }, "size": [ 350, 280 ], "type": "ElevenLabsTextToSpeech", "widgets_values": [ "[low, gravelly voice, unhurried] The lighthouse was dark for eleven years... until tonight.\n\n[quietly, with wonder] I counted the flashes twice, just to be sure. One. Two. Three.\n\n[sharply, alarmed] Wait... that is NOT the pattern we agreed on.\n\n[whispering, urgent] Kill the lamp. We have to go. NOW.\n\n[dry laugh] ...Or we could stay, and see who is climbing the stairs.", 0.5, "auto", "eleven_v4", 0.75, "", 755636914, "randomize", "mp3_44100_192" ], "widgets_values_named": { "apply_text_normalization": "auto", "control_after_generate": "randomize", "language_code": "", "model": "eleven_v4", "model.similarity_boost": 0.75, "output_format": "mp3_44100_192", "seed": 755636914, "stability": 0.5, "text": "[low, gravelly voice, unhurried] The lighthouse was dark for eleven years... until tonight.\n\n[quietly, with wonder] I counted the flashes twice, just to be sure. One. Two. Three.\n\n[sharply, alarmed] Wait... that is NOT the pattern we agreed on.\n\n[whispering, urgent] Kill the lamp. We have to go. NOW.\n\n[dry laugh] ...Or we could stay, and see who is climbing the stairs." } }, { "flags": {}, "id": 210, "inputs": [], "mode": 4, "order": 0, "outputs": [ { "links": [ 296 ], "name": "AUDIO", "type": "AUDIO" } ], "pos": [ 1820, 950 ], "properties": { "Node name for S&R": "LoadAudio", "cnr_id": "comfy-core", "ver": "0.14.1" }, "size": [ 250, 140 ], "type": "LoadAudio", "widgets_values": [ "long_demo.mp3", null ], "widgets_values_named": { "audio": "long_demo.mp3", "upload": null } }, { "flags": {}, "id": 211, "inputs": [], "mode": 0, "order": 1, "outputs": [ { "links": [ 300 ], "name": "voice", "type": "ELEVENLABS_VOICE" } ], "pos": [ 1920, 730 ], "properties": { "Node name for S&R": "ElevenLabsVoiceSelector", "cnr_id": "comfy-core", "ver": "0.14.1" }, "size": [ 320, 80 ], "type": "ElevenLabsVoiceSelector", "widgets_values": [ "Roger (male, american)" ], "widgets_values_named": { "voice": "Roger (male, american)" } }, { "bgcolor": "#653", "color": "#432", "flags": {}, "id": 212, "inputs": [ { "label": "audio0", "link": 296, "name": "files.audio0", "type": "AUDIO" }, { "link": null, "name": "files.audio1", "shape": 7, "type": "AUDIO" } ], "mode": 4, "order": 5, "outputs": [ { "links": null, "name": "voice", "type": "ELEVENLABS_VOICE" } ], "pos": [ 2110, 950 ], "properties": { "Node name for S&R": "ElevenLabsInstantVoiceClone", "cnr_id": "comfy-core", "ver": "0.14.1" }, "size": [ 270, 100 ], "type": "ElevenLabsInstantVoiceClone", "widgets_values": [ true ], "widgets_values_named": { "remove_background_noise": true } }, { "flags": {}, "id": 214, "inputs": [], "mode": 4, "order": 2, "outputs": [ { "links": null, "name": "AUDIO", "type": "AUDIO" } ], "pos": [ 1820, 1180 ], "properties": { "Node name for S&R": "RecordAudio", "cnr_id": "comfy-core", "ver": "0.14.1" }, "size": [ 240, 120 ], "type": "RecordAudio", "widgets_values": [ "", "" ], "widgets_values_named": { "audio": "", "audio#1": "" } }, { "bgcolor": "#000", "color": "#222", "flags": {}, "id": 217, "inputs": [], "mode": 0, "order": 3, "outputs": [], "pos": [ 2120, 1180 ], "properties": {}, "size": [ 260, 100 ], "title": "How to enable bypassed nodes", "type": "Note", "widgets_values": [ "Please select and use Ctrl-B to enable the bypassed nodes." ], "widgets_values_named": { "text": "Please select and use Ctrl-B to enable the bypassed nodes." } }, { "flags": {}, "id": 218, "inputs": [ { "link": 298, "name": "audio", "type": "AUDIO" } ], "mode": 0, "order": 7, "outputs": [ { "links": null, "name": "audio", "type": "AUDIO" } ], "pos": [ 2820, 730 ], "properties": { "Node name for S&R": "SaveAudioAdvanced" }, "size": [ 270, 160 ], "type": "SaveAudioAdvanced", "widgets_values": [ "audio/ComfyUI", "mp3", "V0" ], "widgets_values_named": { "filename_prefix": "audio/ComfyUI", "format": "mp3", "format.quality": "V0" } }, { "bgcolor": "#000", "color": "#222", "flags": {}, "id": 219, "inputs": [], "mode": 0, "order": 4, "outputs": [], "pos": [ 1370, 690 ], "properties": {}, "size": [ 370, 350 ], "type": "MarkdownNote", "widgets_values": [ "## How to write the prompt\n\nEleven v4 and v4 Turbo treat the prompt as a performance script. Audio tags in square brackets set the delivery: `[whispers]`, `[laughs]`, `[sighs]`, `[excited]`, `[sarcastic]`, `[crying]` are a starting point. Place a tag right before or after the words it should affect.\n\n- Describe the voice quality instead of a sound cue: `[low, gravelly voice]` is clearer than a vague tag, because v4 can read an ambiguous tag as a request for a sound effect.\n- Punctuation sets the pacing: ellipses add pauses, capitals add emphasis, standard punctuation keeps a natural rhythm. SSML break tags are not supported, and v4 has no speed or style slider.\n- Match the tags to the voice: a calm narration voice follows `[whispers]` more reliably than `[giggles]`. Combine tags such as `[annoyed] [under your breath]` for a more complex delivery.\n- Stability controls how consistent the delivery stays: lower values are more expressive, higher values more monotone. Similarity controls how closely the output follows the reference voice.\n- `apply_text_normalization` covers numbers, dates and abbreviations. Keep the text within the model's limit: 10,000 characters per request for Eleven v4.\n\nExample prompt:\n\n```text\n[whispers] I never knew it could be this way, but I'm glad we're here.\n```\n\nThe **ElevenLabs Text to Speech** node ships with a demo script that uses these techniques: run it once to hear how the tags land, then replace the text with your own script.\n\nMore techniques: [ElevenLabs prompting guide](https://elevenlabs.io/docs/overview/capabilities/text-to-speech/best-practices#prompting-eleven-v4)\n\n\n## Input Asset\n\n- [long_demo.mp3](https://raw.githubusercontent.com/Comfy-Org/workflow_templates/refs/heads/main/input/long_demo.mp3)\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" ], "widgets_values_named": { "text": "## How to write the prompt\n\nEleven v4 and v4 Turbo treat the prompt as a performance script. Audio tags in square brackets set the delivery: `[whispers]`, `[laughs]`, `[sighs]`, `[excited]`, `[sarcastic]`, `[crying]` are a starting point. Place a tag right before or after the words it should affect.\n\n- Describe the voice quality instead of a sound cue: `[low, gravelly voice]` is clearer than a vague tag, because v4 can read an ambiguous tag as a request for a sound effect.\n- Punctuation sets the pacing: ellipses add pauses, capitals add emphasis, standard punctuation keeps a natural rhythm. SSML break tags are not supported, and v4 has no speed or style slider.\n- Match the tags to the voice: a calm narration voice follows `[whispers]` more reliably than `[giggles]`. Combine tags such as `[annoyed] [under your breath]` for a more complex delivery.\n- Stability controls how consistent the delivery stays: lower values are more expressive, higher values more monotone. Similarity controls how closely the output follows the reference voice.\n- `apply_text_normalization` covers numbers, dates and abbreviations. Keep the text within the model's limit: 10,000 characters per request for Eleven v4.\n\nExample prompt:\n\n```text\n[whispers] I never knew it could be this way, but I'm glad we're here.\n```\n\nThe **ElevenLabs Text to Speech** node ships with a demo script that uses these techniques: run it once to hear how the tags land, then replace the text with your own script.\n\nMore techniques: [ElevenLabs prompting guide](https://elevenlabs.io/docs/overview/capabilities/text-to-speech/best-practices#prompting-eleven-v4)\n\n\n## Input Asset\n\n- [long_demo.mp3](https://raw.githubusercontent.com/Comfy-Org/workflow_templates/refs/heads/main/input/long_demo.mp3)\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" } } ], "revision": 0, "version": 0.4 }