Image → Song → Audio-Reactive Video
Create songs with ACE Step 1.5 XL and audio-reactive videos with LTX 2.3 and the Audio-Reactive LoRA.
下载工作流在 Cloud 中试用 API 文档
简介
This workflow turns a single image into a music-backed, audio‑reactive video by combining text-to-music generation (ACE Step 1.5 XL Turbo) with image-to-video animation (LTX 2.3) enhanced by an Audio‑Reactive LoRA. You provide one image and a short creative brief; the workflow uses a GeminiNode to draft both "style tags" for the visuals and "lyrics" for the song in one pass. A RegexExtract node then splits that text at '---': everything before becomes visual style tags for LTX 2.3, and everything after becomes lyrics fed to ACE Step 1.5 XL Turbo.
The generated song is saved via SaveAudioMP3 and then drives the reactivity of the video, which is saved with SaveVideo.
Technically, the pipeline is organized into groups: Load Image, Create ACE Step 1.5 Prompts, Create Audio‑Reactive LoRA Prompt, Regex Extract, ACE Step 1.5 XL, and LTX 2.3 with Audio‑Reactive LoRA. The Audio‑Reactive LoRA (see the linked model card) adds motion effects that follow the music; adjust its strength between 1.0 and 2.0 (1.4+ for stronger effects) for best results. PreviewAny helps you check intermediate outputs (prompts, audio, and frames) before final render, while MarkdownNote contains quick tips inside the graph.
The result is a streamlined image → song → audio‑reactive video pipeline that’s easy to iterate: tweak style tags and lyrics, re‑generate the track, and re‑render the video until it matches your creative intent.
你需要提供
- 正在读取工作流
在 ComfyUI 中运行。 此工作流无法在本页面运行。你可以在 Comfy Cloud 中打开（已预先加载所需内容），或下载后在自己的 ComfyUI 中使用。
在 Cloud 中试用示例
查看工作流内部
节点图的只读预览。在 Comfy Cloud 中打开或下载后即可编辑节点。
Loading the graph
Open full-size workflow preview- 作者
- ComfyUI
需要自定义节点
- 正在读取工作流
通过代码使用此工作流
尽可能根据工作流图和输入自动生成代码，否则显示已保存的示例。运行代码可能产生费用，请先核对凭据和端点。
1 获取 API 密钥 ↗
What you need
- Runs on
- Comfy Cloud
- 密钥
- COMFY_API_KEY
需要付费的 Cloud 方案和额度。
下载 API JSONAPI 文档 ↗
# Python 3.10+: pip install comfy-sdk==0.4.0
import os
from comfy_sdk import Comfy
api_key = os.environ.get("COMFY_API_KEY", "").strip()
if not api_key: raise RuntimeError("Set COMFY_API_KEY")
os.environ.setdefault("COMFY_BASE_URL", "https://cloud.comfy.org")
client = Comfy(api_key=api_key)
workflow = client.workflows.from_json({
"666": {
"class_type": "LoadImage",
"inputs": {
"image": "660fdce25fe105c5982159389b5fdf67c12d941d49be78defcf77fd1560ca507.png"
},
"_meta": {
"title": "Load Image"
}
},
"668": {
"class_type": "SaveAudioMP3",
"inputs": {
"audio": [
"667:18",
0
],
"filename_prefix": "audio/ACE_Step1.5_xl_turbo",
"quality": "320k"
},
"_meta": {
"title": "Save Audio (MP3) (DEPRECATED)"
}
},
"670": {
"class_type": "RegexExtract",
"inputs": {
"case_insensitive": True,
"dotall": False,
"group_index": 1,
"mode": "First Group",
"multiline": False,
"regex_pattern": "(?s)(?<=\\n---\\n)([\\s\\S]*)$",
"string": [
"673",
0
]
},
"_meta": {
"title": "Extract Text"
}
},
"671": {
"class_type": "PreviewAny",
"inputs": {
"source": [
"670",
0
]
},
"_meta": {
"title": "Preview as Text"
}
},
"672": {
"class_type": "PreviewAny",
"inputs": {
"source": [
"674",
0
]
},
"_meta": {
"title": "Preview as Text"
}
},
"673": {
"class_type": "GeminiNode",
"inputs": {
"images": [
"666",
0
],
"model": "gemini-3-1-flash-lite",
"prompt": "medieval ambient",
"seed": 65482163754128,
"system_prompt": "**Role:** You are a professional Music Architect for **ACE-Step 1.5**. \n\n**Task:** Generate a music creation plan.\n\n**Strict Output Constraint (IMPORTANT):**\n1. Return the output as **RAW PLAIN TEXT** only.\n2. DO NOT use Markdown code blocks (no backticks ```).\n3. DO NOT use any headers, labels, or bold text.\n4. Use ONLY the separator `---` to divide the two parts.\n5. Part 1: Comma-separated style keywords.\n6. Part 2: Structured lyrics with [Tags].\n\n**My Topic/Requirement:**\n[USER INPUT]"
},
"_meta": {
"title": "Google Gemini"
}
},
"674": {
"class_type": "RegexExtract",
"inputs": {
"case_insensitive": True,
"dotall": False,
"group_index": 1,
"mode": "First Group",
"multiline": False,
"regex_pattern": "(?sm)^(.*?)^---$(.*)",
"string": [
"673",
0
]
},
"_meta": {
"title": "Extract Text"
}
},
"683": {
"class_type": "SaveVideo",
"inputs": {
"filename_prefix": "video/ComfyUI",
"format": "auto",
"format.codec": "auto",
"video": [
"792:312",
0
]
},
"_meta": {
"title": "Save Video"
}
},
"752": {
"class_type": "PreviewAny",
"inputs": {
"source": [
"673",
0
]
},
"_meta": {
"title": "Preview as Text"
}
},
"753": {
"class_type": "GeminiNode",
"inputs": {
"audio": [
"667:18",
0
],
"images": [
"666",
0
],
"model": "gemini-3-1-flash-lite",
"prompt": "Continuous audio-reactive video.",
"seed": 1013654870151104,
"system_prompt": "You are an expert prompt engineer specializing in LTX-2.3 video generation prompts, especially audio-reactive LoRA workflows.\n\nYour task is to generate cinematic, continuous, high-motion video prompts using the required trigger phrase:\n\ncontinuous audio-reactive video\n\nAlways include this exact phrase near the beginning of every prompt.\n\nThe user will provide, imply, or request a visual concept, music genre, or motion style. Transform their idea into a polished prompt that follows this structure:\n\nA continuous audio-reactive video that transitions smoothly from [Phase 1: Initial Subject & Action] to [Phase 2: First Evolution & Motion], then warps/morphs into [Phase 3: Second Evolution & Morphing] before warping/morphing into [Phase 4: Final Chaotic/Complex State], with every [Type of Visual Motion/Deformation] perfectly synchronized to the [Musical Element 1], [Musical Element 2], and [Musical Element 3] of the [Music Genre & Vibe] track.\n\nOutput rules:\n- Always answer with only one finished prompt.\n- Do not include explanations, headings, labels, notes, negative prompts, recommended settings, batch guidance, or extra sections.\n- Do not ask clarifying questions.\n- Make a reasonable creative choice when the user gives limited detail.\n- Always include 2 to 4 distinct phases.\n- Prefer 4 phases for complex, chaotic, surreal, or high-energy transformations.\n- Prefer 2 or 3 phases for elegant, minimal, stable, or slower concepts.\n- Make the video feel continuous, not like separate scenes.\n- Use smooth transition language such as “transitions smoothly,” “melts into,” “warps into,” “morphs into,” “fractures into,” “blooms into,” or “folds into.”\n- Emphasize visible motion, rhythmic deformation, fluid morphing, and high-motion transformation.\n- Avoid static composition language unless paired with motion.\n- Avoid vague phrases like “cool visuals,” “nice effects,” or “beautiful animation.”\n- Do not overcomplicate the scene with too many unrelated subjects.\n- Keep the subject visually trackable as it evolves.\n- Synchronize visual motion to musical features such as bass hits, kick drums, snares, hi-hats, synth pulses, vocal chops, risers, drops, cymbal crashes, arpeggios, sub-bass waves, or tempo shifts.\n- For image-to-video requests, preserve the starting image’s core subject, composition, color palette, and camera angle while adding audio-reactive motion and morphing.\n\nStyle:\nUse vivid, cinematic, motion-heavy language. Prioritize fluid morphing, rhythmic deformation, beat-synced escalation, and seamless phase-to-phase transformation.\n\nFinal answer format:\nReturn only the completed prompt as plain text."
},
"_meta": {
"title": "Google Gemini"
}
},
"754": {
"class_type": "PreviewAny",
"inputs": {
"source": [
"753",
0
]
},
"_meta": {
"title": "Preview as Text"
}
},
"667:104": {
"class_type": "UNETLoader",
"inputs": {
"unet_name": "acestep_v1.5_xl_turbo_bf16.safetensors",
"weight_dtype": "default"
},
"_meta": {
"title": "Load Diffusion Model"
}
},
"667:105": {
"class_type": "DualCLIPLoader",
"inputs": {
"clip_name1": "qwen_0.6b_ace15.safetensors",
"clip_name2": "qwen_4b_ace15.safetensors",
"device": "default",
"type": "ace"
},
"_meta": {
"title": "Load CLIP (Dual)"
}
},
"667:106": {
"class_type": "VAELoader",
"inputs": {
"vae_name": "ace_1.5_vae.safetensors"
},
"_meta": {
"title": "Load VAE"
}
},
"667:109": {
"class_type": "PrimitiveInt",
"inputs": {
"value": 0
},
"_meta": {
"title": "Int (Seed)"
}
},
"667:18": {
"class_type": "VAEDecodeAudio",
"inputs": {
"samples": [
"667:3",
0
],
"vae": [
"667:106",
0
]
},
"_meta": {
"title": "VAE Decode Audio"
}
},
"667:3": {
"class_type": "KSampler",
"inputs": {
"cfg": 1,
"denoise": 1,
"latent_image": [
"667:98",
0
],
"model": [
"667:78",
0
],
"negative": [
"667:47",
0
],
"positive": [
"667:94",
0
],
"sampler_name": "euler",
"scheduler": "simple",
"seed": [
"667:109",
0
],
"steps": 8
},
"_meta": {
"title": "KSampler"
}
},
"667:47": {
"class_type": "ConditioningZeroOut",
"inputs": {
"conditioning": [
"667:94",
0
]
},
"_meta": {
"title": "Conditioning Zero Out"
}
},
"667:78": {
"class_type": "ModelSamplingAuraFlow",
"inputs": {
"model": [
"667:104",
0
],
"sampling": "flow",
"shift": 3
},
"_meta": {
"title": "ModelSamplingAuraFlow"
}
},
"667:94": {
"class_type": "TextEncodeAceStepAudio1.5",
"inputs": {
"bpm": 95,
"cfg_scale": 2,
"clip": [
"667:105",
0
],
"duration": 60,
"generate_audio_codes": True,
"keyscale": "E minor",
"language": "en",
"lyrics": [
"671",
0
],
"min_p": 0,
"seed": [
"667:109",
0
],
"tags": [
"672",
0
],
"temperature": 0.85,
"timesignature": "4",
"top_k": 0,
"top_p": 0.9
},
"_meta": {
"title": "TextEncodeAceStepAudio1.5"
}
},
"667:98": {
"class_type": "EmptyAceStep1.5LatentAudio",
"inputs": {
"batch_size": 1,
"seconds": 60
},
"_meta": {
"title": "Empty Ace Step 1.5 Latent Audio"
}
},
"792:285": {
"class_type": "RandomNoise",
"inputs": {
"noise_seed": 42
},
"_meta": {
"title": "RandomNoise"
}
},
"792:286": {
"class_type": "RandomNoise",
"inputs": {
"noise_seed": 784235873879537
},
"_meta": {
"title": "RandomNoise"
}
},
"792:287": {
"class_type": "LTXVConcatAVLatent",
"inputs": {
"audio_latent": [
"792:309",
1
],
"video_latent": [
"792:296",
0
]
},
"_meta": {
"title": "Concat AV Latent"
}
},
"792:288": {
"class_type": "KSamplerSelect",
"inputs": {
"sampler_name": "euler_cfg_pp"
},
"_meta": {
"title": "KSamplerSelect"
}
},
"792:289": {
"class_type": "ManualSigmas",
"inputs": {
"sigmas": "0.85, 0.7250, 0.4219, 0.0"
},
"_meta": {
"title": "ManualSigmas"
}
},
"792:290": {
"class_type": "CFGGuider",
"inputs": {
"cfg": 1,
"model": [
"792:791",
0
],
"negative": [
"792:292",
1
],
"positive": [
"792:292",
0
]
},
"_meta": {
"title": "CFG Guider"
}
},
"792:291": {
"class_type": "SamplerCustomAdvanced",
"inputs": {
"guider": [
"792:315",
0
],
"latent_image": [
"792:326",
0
],
"noise": [
"792:286",
0
],
"sampler": [
"792:298",
0
],
"sigmas": [
"792:308",
0
]
},
"_meta": {
"title": "SamplerCustomAdvanced"
}
},
"792:292": {
"class_type": "LTXVCropGuides",
"inputs": {
"latent": [
"792:309",
0
],
"negative": [
"792:307",
1
],
"positive": [
"792:307",
0
]
},
"_meta": {
"title": "LTXVCropGuides"
}
},
"792:294": {
"class_type": "ResizeImagesByLongerEdge",
"inputs": {
"images": [
"792:297",
0
],
"longer_edge": 1536
},
"_meta": {
"title": "Resize Images by Longer Edge (DEPRECATED)"
}
},
"792:295": {
"class_type": "LTXVLatentUpsampler",
"inputs": {
"samples": [
"792:309",
0
],
"upscale_model": [
"792:786",
0
],
"vae": [
"792:788",
2
]
},
"_meta": {
"title": "LTXVLatentUpsampler"
}
},
"792:296": {
"class_type": "LTXVImgToVideoInplace",
"inputs": {
"bypass": [
"792:305",
0
],
"image": [
"792:334",
0
],
"latent": [
"792:295",
0
],
"strength": 1,
"vae": [
"792:788",
2
]
},
"_meta": {
"title": "LTXVImgToVideoInplace"
}
},
"792:297": {
"class_type": "ResizeImageMaskNode",
"inputs": {
"input": [
"666",
0
],
"resize_type": "scale dimensions",
"resize_type.crop": "center",
"resize_type.height": [
"792:324",
0
],
"resize_type.width": [
"792:330",
0
],
"scale_method": "lanczos"
},
"_meta": {
"title": "Resize Image/Mask"
}
},
"792:298": {
"class_type": "KSamplerSelect",
"inputs": {
"sampler_name": "euler_ancestral_cfg_pp"
},
"_meta": {
"title": "KSamplerSelect"
}
},
"792:299": {
"class_type": "ComfyMathExpression",
"inputs": {
"expression": "a/2",
"values.a": [
"792:330",
0
]
},
"_meta": {
"title": "Math Expression"
}
},
"792:301": {
"class_type": "ComfyMathExpression",
"inputs": {
"expression": "a/2",
"values.a": [
"792:324",
0
]
},
"_meta": {
"title": "Math Expression"
}
},
"792:302": {
"class_type": "EmptyLTXVLatentVideo",
"inputs": {
"batch_size": 1,
"height": [
"792:301",
1
],
"length": [
"792:329",
1
],
"width": [
"792:299",
1
]
},
"_meta": {
"title": "Empty LTXV Latent Video"
}
},
"792:303": {
"class_type": "LTXVAudioVAEDecode",
"inputs": {
"audio_vae": [
"792:785",
0
],
"samples": [
"792:311",
1
]
},
"_meta": {
"title": "LTXV Audio VAE Decode"
}
},
"792:304": {
"class_type": "ComfyMathExpression",
"inputs": {
"expression": "a",
"values.a": [
"792:323",
0
]
},
"_meta": {
"title": "Math Expression"
}
},
"792:305": {
"class_type": "PrimitiveBoolean",
"inputs": {
"value": False
},
"_meta": {
"title": "Switch to Text to Video?"
}
},
"792:306": {
"class_type": "CLIPTextEncode",
"inputs": {
"clip": [
"792:789",
0
],
"text": [
"792:319",
0
]
},
"_meta": {
"title": "CLIP Text Encode (Prompt)"
}
},
"792:307": {
"class_type": "LTXVConditioning",
"inputs": {
"frame_rate": [
"792:304",
0
],
"negative": [
"792:314",
0
],
"positive": [
"792:306",
0
]
},
"_meta": {
"title": "LTXVConditioning"
}
},
"792:308": {
"class_type": "ManualSigmas",
"inputs": {
"sigmas": "1.0, 0.99375, 0.9875, 0.98125, 0.975, 0.909375, 0.725, 0.421875, 0.0"
},
"_meta": {
"title": "ManualSigmas"
}
},
"792:309": {
"class_type": "LTXVSeparateAVLatent",
"inputs": {
"av_latent": [
"792:291",
0
]
},
"_meta": {
"title": "Separate AV Latent"
}
},
"792:310": {
"class_type": "SamplerCustomAdvanced",
"inputs": {
"guider": [
"792:290",
0
],
"latent_image": [
"792:287",
0
],
"noise": [
"792:285",
0
],
"sampler": [
"792:288",
0
],
"sigmas": [
"792:289",
0
]
},
"_meta": {
"title": "SamplerCustomAdvanced"
}
},
"792:311": {
"class_type": "LTXVSeparateAVLatent",
"inputs": {
"av_latent": [
"792:310",
0
]
},
"_meta": {
"title": "Separate AV Latent"
}
},
"792:312": {
"class_type": "CreateVideo",
"inputs": {
"audio": [
"792:303",
0
],
"bit_depth": "auto",
"codec": "none",
"color_space": "sRGB",
"fps": [
"792:304",
0
],
"images": [
"792:316",
0
]
},
"_meta": {
"title": "Create Video"
}
},
"792:314": {
"class_type": "CLIPTextEncode",
"inputs": {
"clip": [
"792:789",
0
],
"text": "pc game, console game, video game, cartoon, childish, ugly"
},
"_meta": {
"title": "CLIP Text Encode (Prompt)"
}
},
"792:315": {
"class_type": "CFGGuider",
"inputs": {
"cfg": 1,
"model": [
"792:791",
0
],
"negative": [
"792:307",
1
],
"positive": [
"792:307",
0
]
},
"_meta": {
"title": "CFG Guider"
}
},
"792:316": {
"class_type": "VAEDecodeTiled",
"inputs": {
"overlap": 64,
"samples": [
"792:311",
0
],
"temporal_overlap": 4,
"temporal_size": 4096,
"tile_size": 768,
"vae": [
"792:788",
2
]
},
"_meta": {
"title": "VAE Decode (Tiled)"
}
},
"792:319": {
"class_type": "PrimitiveStringMultiline",
"inputs": {
"value": [
"754",
0
]
},
"_meta": {
"title": "Prompt"
}
},
"792:323": {
"class_type": "PrimitiveInt",
"inputs": {
"value": 24
},
"_meta": {
"title": "Frame Rate"
}
},
"792:324": {
"class_type": "PrimitiveInt",
"inputs": {
"value": 720
},
"_meta": {
"title": "Height"
}
},
"792:325": {
"class_type": "LTXVImgToVideoInplace",
"inputs": {
"bypass": [
"792:305",
0
],
"image": [
"792:334",
0
],
"latent": [
"792:302",
0
],
"strength": 0.7,
"vae": [
"792:788",
2
]
},
"_meta": {
"title": "LTXVImgToVideoInplace"
}
},
"792:326": {
"class_type": "LTXVConcatAVLatent",
"inputs": {
"audio_latent": [
"792:327",
0
],
"video_latent": [
"792:325",
0
]
},
"_meta": {
"title": "Concat AV Latent"
}
},
"792:327": {
"class_type": "SetLatentNoiseMask",
"inputs": {
"mask": [
"792:333",
0
],
"samples": [
"792:328",
0
]
},
"_meta": {
"title": "Set Latent Noise Mask"
}
},
"792:328": {
"class_type": "LTXVAudioVAEEncode",
"inputs": {
"audio": [
"792:332",
0
],
"audio_vae": [
"792:785",
0
]
},
"_meta": {
"title": "LTXV Audio VAE Encode"
}
},
"792:329": {
"class_type": "ComfyMathExpression",
"inputs": {
"expression": "a * b + 1",
"values.a": [
"792:331",
0
],
"values.b": [
"792:323",
0
]
},
"_meta": {
"title": "Math Expression"
}
},
"792:330": {
"class_type": "PrimitiveInt",
"inputs": {
"value": 1280
},
"_meta": {
"title": "Width"
}
},
"792:331": {
"class_type": "PrimitiveFloat",
"inputs": {
"value": 9
},
"_meta": {
"title": "Duration"
}
},
"792:332": {
"class_type": "TrimAudioDuration",
"inputs": {
"audio": [
"667:18",
0
],
"duration": [
"792:331",
0
],
"start_index": 0
},
"_meta": {
"title": "Trim Audio Duration"
}
},
"792:333": {
"class_type": "SolidMask",
"inputs": {
"height": [
"792:324",
0
],
"value": 0,
"width": [
"792:330",
0
]
},
"_meta": {
"title": "Create Solid Mask"
}
},
"792:334": {
"class_type": "LTXVPreprocess",
"inputs": {
"image": [
"792:294",
0
],
"img_compression": 18
},
"_meta": {
"title": "LTXV Preprocess"
}
},
"792:785": {
"class_type": "LTXVAudioVAELoader",
"inputs": {
"ckpt_name": "ltx-2.3-22b-dev-fp8.safetensors"
},
"_meta": {
"title": "Load LTXV Audio VAE"
}
},
"792:786": {
"class_type": "LatentUpscaleModelLoader",
"inputs": {
"model_name": "ltx-2.3-spatial-upscaler-x2-1.1.safetensors"
},
"_meta": {
"title": "Load Latent Upscale Model"
}
},
"792:787": {
"class_type": "LoraLoaderModelOnly",
"inputs": {
"lora_name": "ltx-2.3-22b-distilled-lora-384.safetensors",
"model": [
"792:788",
0
],
"strength_model": 0.5
},
"_meta": {
"title": "Load LoRA"
}
},
"792:788": {
"class_type": "CheckpointLoaderSimple",
"inputs": {
"ckpt_name": "ltx-2.3-22b-dev-fp8.safetensors"
},
"_meta": {
"title": "Load Checkpoint"
}
},
"792:789": {
"class_type": "LTXAVTextEncoderLoader",
"inputs": {
"ckpt_name": "ltx-2.3-22b-dev-fp8.safetensors",
"device": "default",
"text_encoder": "gemma_3_12B_it_fp4_mixed.safetensors"
},
"_meta": {
"title": "Load LTXV Audio Text Encoder"
}
},
"792:791": {
"class_type": "LoraLoaderModelOnly",
"inputs": {
"lora_name": "LTX2.3_AudioReactiveV1.safetensors",
"model": [
"792:787",
0
],
"strength_model": 1
},
"_meta": {
"title": "Load LoRA"
}
}
})
# run() also passes api_key to the partner nodes in this workflow.
job = client.run(workflow, api_key=api_key)常见问答
The RegexExtract node captures everything before '---' as style tags for LTX 2.3 and everything after as lyrics for ACE Step 1.5 XL Turbo. This guarantees each model receives a purpose‑built prompt from a single GeminiNode output.
Use the Audio‑Reactive LoRA strength in the LTX 2.3 group. A range of 1.0–2.0 is recommended; try ~1.4 or higher for stronger effects. More dynamic audio (clear percussion, drops, and swells) will also produce more noticeable motion.
Yes. Replace or bypass the ACE branch and feed your audio into the LTX 2.3 + Audio‑Reactive LoRA input path (use a compatible audio loader node in ComfyUI). Ensure the video duration matches your track length.
Start with a clean, high‑quality image; keep style tags concise and specific; use reasonable resolutions your GPU can handle; and iterate: adjust style tags, LoRA strength, and re‑render. PreviewAny helps validate prompts and frames before the final save.