混元影片 1.5 圖生影片
將靜態圖像動畫化為動態影片,支援精準的運動與鏡頭控制,在保持照片與插畫視覺一致的同時,帶來流暢自然的動作,讓畫面更生動。
工作流程內部
來源範本的唯讀預覽。請在 Cloud 中開啟,以探索節點並進行變更。
從您的程式碼呼叫此模型
以下請求使用與 Playground 相同的已準備節點圖和設定。每次提交都可能啟動付費執行;若提交結果不明,請勿自動重試提交。
What you need
- Runs on
- Comfy Cloud
- 金鑰
- COMFY_API_KEY
- 您的檔案
- 由程式碼透過其 URL 上傳
需要付費 Cloud 方案與點數。
# 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({
"8": {
"class_type": "VAEDecode",
"inputs": {
"samples": [
"125",
0
],
"vae": [
"10",
0
]
},
"_meta": {
"title": "VAE Decode"
}
},
"10": {
"class_type": "VAELoader",
"inputs": {
"vae_name": "hunyuanvideo15_vae_fp16.safetensors"
},
"_meta": {
"title": "Load VAE"
}
},
"11": {
"class_type": "DualCLIPLoader",
"inputs": {
"clip_name1": "qwen_2.5_vl_7b_fp8_scaled.safetensors",
"clip_name2": "byt5_small_glyphxl_fp16.safetensors",
"device": "default",
"type": "hunyuan_video_15"
},
"_meta": {
"title": "Load CLIP (Dual)"
}
},
"12": {
"class_type": "UNETLoader",
"inputs": {
"unet_name": "hunyuanvideo1.5_720p_i2v_fp16.safetensors",
"weight_dtype": "default"
},
"_meta": {
"title": "Load Diffusion Model"
}
},
"44": {
"class_type": "CLIPTextEncode",
"inputs": {
"clip": [
"11",
0
],
"text": "The bull elk stands on the mossy granite ledge above the fog-filled valley at dusk. It lifts its heavy antlered head, exhales a long plume of breath that curls into the cold air, then turns its neck to look back over its shoulder toward the camera before stepping carefully forward along the rock edge, hooves slipping slightly on the wet stone. Below, thick pewter mist rolls slowly through the dark hemlocks and spills over the ridges like a slow tide. The camera starts close at the elk's shoulder and cranes up and out in one smooth move, rising above the ledge to reveal the full misty valley stretching away, with silver rim light flickering through breaks in the storm clouds. Cinematic wildlife footage, natural motion, steady and unhurried."
},
"_meta": {
"title": "CLIP Text Encode (Positive Prompt)"
}
},
"78": {
"class_type": "HunyuanVideo15ImageToVideo",
"inputs": {
"batch_size": 1,
"clip_vision_output": [
"79",
0
],
"height": 720,
"length": 121,
"negative": [
"93",
0
],
"positive": [
"44",
0
],
"start_image": [
"80",
0
],
"vae": [
"10",
0
],
"width": 1280
},
"_meta": {
"title": "HunyuanVideo15ImageToVideo"
}
},
"79": {
"class_type": "CLIPVisionEncode",
"inputs": {
"clip_vision": [
"81",
0
],
"crop": "center",
"image": [
"80",
0
]
},
"_meta": {
"title": "CLIP Vision Encode"
}
},
"80": {
"class_type": "LoadImage",
"inputs": {
"image": "UPLOADED_image_FILENAME"
},
"_meta": {
"title": "Load Image"
}
},
"81": {
"class_type": "CLIPVisionLoader",
"inputs": {
"clip_name": "sigclip_vision_patch14_384.safetensors"
},
"_meta": {
"title": "Load CLIP Vision"
}
},
"93": {
"class_type": "CLIPTextEncode",
"inputs": {
"clip": [
"11",
0
],
"text": ""
},
"_meta": {
"title": "CLIP Text Encode (Negative Prompt)"
}
},
"101": {
"class_type": "CreateVideo",
"inputs": {
"bit_depth": "auto",
"codec": "none",
"color_space": "sRGB",
"fps": 24,
"images": [
"8",
0
]
},
"_meta": {
"title": "Create Video"
}
},
"102": {
"class_type": "SaveVideo",
"inputs": {
"filename_prefix": "video/hunyuan_video_1.5",
"format": "auto",
"format.codec": "h264",
"video": [
"101",
0
]
},
"_meta": {
"title": "Save Video"
}
},
"125": {
"class_type": "SamplerCustomAdvanced",
"inputs": {
"guider": [
"129",
0
],
"latent_image": [
"78",
2
],
"noise": [
"127",
0
],
"sampler": [
"128",
0
],
"sigmas": [
"126",
0
]
},
"_meta": {
"title": "SamplerCustomAdvanced"
}
},
"126": {
"class_type": "BasicScheduler",
"inputs": {
"denoise": 1,
"model": [
"12",
0
],
"scheduler": "simple",
"steps": 20
},
"_meta": {
"title": "BasicScheduler"
}
},
"127": {
"class_type": "RandomNoise",
"inputs": {
"noise_seed": 887963123424675
},
"_meta": {
"title": "RandomNoise"
}
},
"128": {
"class_type": "KSamplerSelect",
"inputs": {
"sampler_name": "euler"
},
"_meta": {
"title": "KSamplerSelect"
}
},
"129": {
"class_type": "CFGGuider",
"inputs": {
"cfg": 6,
"model": [
"130",
0
],
"negative": [
"78",
1
],
"positive": [
"78",
0
]
},
"_meta": {
"title": "CFG Guider"
}
},
"130": {
"class_type": "ModelSamplingSD3",
"inputs": {
"model": [
"12",
0
],
"shift": 7
},
"_meta": {
"title": "ModelSamplingSD3"
}
}
})
workflow.set_input("80", "image", client.assets.from_url("https://comfy.org/catalog-media/d312a006-8814-4525-9416-61b1de7a21ed/bea0b5d9-b558-48de-9899-f526d73b354f"))
# run() also passes api_key to the partner nodes in this workflow.
job = client.run(workflow, api_key=api_key)
for output in job.get_outputs("102"):
output.to_file(output.name)試試範例
關於此工作流程
Hunyuan Video 1.5 圖像轉影片工作流程,專為將靜態圖片轉換為具動態感的影片而設計,並可精確控制動作與鏡頭。此流程運用 Hunyuan Video 模型,能在維持視覺一致性的同時,為圖片帶來流暢自然的動態。主要節點包含 VAEDecode、VAELoader 與 HunyuanVideo15ImageToVideo,協同處理圖片解碼與影片生成。流程也結合 CLIPTextEncode 與 CLIPVisionEncode 等進階節點,提升語意理解與視覺編碼,確保生成影片既連貫又具吸引力。
技術上,流程分為模型載入、圖片上傳、提示詞設定與影片尺寸調整等步驟。利用 CreateVideo 與 SaveVideo 節點,確保最終輸出高效且正確儲存。EasyCache 節點可選擇性加速處理,但會犧牲部分畫質,讓使用者可依硬體資源彈性調整。非常適合藝術家、動畫師與內容創作者,輕鬆將靜態圖像賦予生命。
問答
EasyCache 節點可加速流程運算,但會降低影片品質。若硬體資源有限時可考慮啟用。
可將推理步數提升至 20 步以上,或在不考慮處理時間時選用 VAE Decode(Tiled)選項。
可將 'weight_dtype' 設為 'fp8_e4n3fn' 以降低記憶體用量,或改用 VRAM 較大的機器。
可以,啟用相關選項即可將影片放大至 1080P,但會增加運算需求。
相關工作流程
- Hunyuan Video
混元影片 1.5 文生影片
- Hunyuan Video, Tencent
混元影片文生影片
- LTX-2.3
Turn an image into a video
- MiniMax H3
Create a video from references
- LTX-2.3
Add camera motion first last frame
- Wan2.2
1.2 入門範本-圖像轉影片
- 合作夥伴節點
AI on the Lot 2026 - LTX-2.3 - i2v
- Google Gemini, Kling 3.0, Z-Image-Turbo
AIrt MAchIne(API 版本)
- SSD1.5
AnimateDiff - IPAdapters 迴圈動畫
- Seedance1.0 Pro, ByteDance
ByteDance: 起始結束到影片
- 合合作夥伴節點
FastVideo FastH3: Text to Video
- Flux
FLUX 3 影片:文字轉影片