AIrt MAchIne (API version)
Upload an image to generate a unique, looping AI video. The workflow uses an LLM to create prompts, employs Kling3.0 for video generation, and stitches the clips into a seamless final loop.
Download WorkflowTry in Cloud API documentation
About
The AIrt MAchIne workflow is designed to transform a static image into a dynamic, looping AI-generated video. This innovative process begins with an image upload, which serves as the foundation for generating a unique video sequence. The workflow leverages a Large Language Model (LLM) to create prompts that guide the video generation process, ensuring each output is distinct and creative. The Kling 3.0 model is central to video creation, providing advanced capabilities for generating high-quality, seamless video clips.
These clips are then stitched together to form a continuous loop, offering a visually engaging experience. The workflow utilizes a combination of nodes, including ImageResizeKJv2 for resizing images, GeminiNode for integrating Google's Gemini model, and VHS_VideoCombine for merging video components. This setup allows users to produce professional-quality videos with minimal input, making it a valuable tool for content creators and marketers looking to enhance their visual media with AI-driven creativity.
What you provide
- Reading the workflow
Runs in ComfyUI. This workflow can't run on this page. Open it in Comfy Cloud with everything loaded, or download it for your own ComfyUI.
Try in CloudExample
Inside the workflow
A read-only preview of the node graph. Open it in Comfy Cloud or download it to edit the nodes.
Loading the graph
Open full-size workflow preview- Author
- Julien Durand | MJM
Requires custom nodes
- Reading the workflow
Use this workflow from code
Generated from the workflow graph and inputs when possible; stored examples are shown otherwise. Review credentials and endpoints before running code, which may incur charges.
1 Get API key ↗
What you need
- Runs on
- Comfy Cloud
- Key
- COMFY_API_KEY
Needs a paid Cloud plan and credits.
Download API JSONAPI documentation ↗
# 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({
"5": {
"class_type": "ImageResizeKJv2",
"inputs": {
"crop_position": "center",
"device": "cpu",
"divisible_by": 1,
"height": [
"163",
0
],
"image": [
"99:77",
0
],
"keep_proportion": "crop",
"pad_color": "0, 0, 0",
"upscale_method": "lanczos",
"width": [
"162",
0
]
},
"_meta": {
"title": "Resize Image v2"
}
},
"9": {
"class_type": "ImageResizeKJv2",
"inputs": {
"crop_position": "center",
"device": "cpu",
"divisible_by": 1,
"height": [
"163",
0
],
"image": [
"103:201",
0
],
"keep_proportion": "crop",
"pad_color": "0, 0, 0",
"upscale_method": "lanczos",
"width": [
"162",
0
]
},
"_meta": {
"title": "Resize Image v2"
}
},
"10": {
"class_type": "SaveVideo",
"inputs": {
"filename_prefix": "MJM/AIrt MAchIne/Clips/MJM_AIrtMAchIne_Vid1",
"format": "auto",
"format.codec": "auto",
"video": [
"139:247",
0
]
},
"_meta": {
"title": "Save Video"
}
},
"14": {
"class_type": "ImageResizeKJv2",
"inputs": {
"crop_position": "center",
"device": "cpu",
"divisible_by": 1,
"height": [
"163",
0
],
"image": [
"107:213",
0
],
"keep_proportion": "crop",
"pad_color": "0, 0, 0",
"upscale_method": "lanczos",
"width": [
"162",
0
]
},
"_meta": {
"title": "Resize Image v2"
}
},
"16": {
"class_type": "SaveVideo",
"inputs": {
"filename_prefix": "MJM/AIrt MAchIne/Clips/MJM_AIrtMAchIne_Vid2",
"format": "auto",
"format.codec": "auto",
"video": [
"141:250",
0
]
},
"_meta": {
"title": "Save Video"
}
},
"20": {
"class_type": "ImageResizeKJv2",
"inputs": {
"crop_position": "center",
"device": "cpu",
"divisible_by": 1,
"height": [
"163",
0
],
"image": [
"111:229",
0
],
"keep_proportion": "crop",
"pad_color": "0, 0, 0",
"upscale_method": "lanczos",
"width": [
"162",
0
]
},
"_meta": {
"title": "Resize Image v2"
}
},
"22": {
"class_type": "SaveVideo",
"inputs": {
"filename_prefix": "MJM/AIrt MAchIne/Clips/MJM_AIrtMAchIne_Vid3",
"format": "auto",
"format.codec": "auto",
"video": [
"144:253",
0
]
},
"_meta": {
"title": "Save Video"
}
},
"25": {
"class_type": "ImageResizeKJv2",
"inputs": {
"crop_position": "center",
"device": "cpu",
"divisible_by": 1,
"height": [
"163",
0
],
"image": [
"115:240",
0
],
"keep_proportion": "crop",
"pad_color": "0, 0, 0",
"upscale_method": "lanczos",
"width": [
"162",
0
]
},
"_meta": {
"title": "Resize Image v2"
}
},
"27": {
"class_type": "SaveVideo",
"inputs": {
"filename_prefix": "MJM/AIrt MAchIne/Clips/MJM_AIrtMAchIne_Vid4",
"format": "auto",
"format.codec": "auto",
"video": [
"146:256",
0
]
},
"_meta": {
"title": "Save Video"
}
},
"33": {
"class_type": "SaveVideo",
"inputs": {
"filename_prefix": "MJM/AIrt MAchIne/Clips/MJM_AIrtMAchIne_Vid5",
"format": "auto",
"format.codec": "auto",
"video": [
"148:259",
0
]
},
"_meta": {
"title": "Save Video"
}
},
"94": {
"class_type": "LoadImage",
"inputs": {
"image": "AIrt_MAchIne_init.png"
},
"_meta": {
"title": "Load Image"
}
},
"95": {
"class_type": "GeminiNode",
"inputs": {
"images": [
"94",
0
],
"model": "gemini-3-pro-preview",
"prompt": "write 5 prompts based on this image, I will use these in a text 2 image model. The prompts should be different, be creative. Only use the image as inspiration.\n\nI want only the prompts and nothing else. Each prompt on its own line.\n\nNo empty lines in-between the prompts:\nPrompt 1\nPrompt 2\nPrompt 3\nPrompt 4\nPrompt 5",
"seed": 121362321499456,
"system_prompt": ""
},
"_meta": {
"title": "Google Gemini"
}
},
"100": {
"class_type": "PreviewImage",
"inputs": {
"images": [
"99:73",
0
]
},
"_meta": {
"title": "Preview Image"
}
},
"104": {
"class_type": "PreviewImage",
"inputs": {
"images": [
"103:200",
0
]
},
"_meta": {
"title": "Preview Image"
}
},
"108": {
"class_type": "PreviewImage",
"inputs": {
"images": [
"107:217",
0
]
},
"_meta": {
"title": "Preview Image"
}
},
"112": {
"class_type": "PreviewImage",
"inputs": {
"images": [
"111:226",
0
]
},
"_meta": {
"title": "Preview Image"
}
},
"116": {
"class_type": "PreviewImage",
"inputs": {
"images": [
"115:239",
0
]
},
"_meta": {
"title": "Preview Image"
}
},
"119": {
"class_type": "GetVideoComponents",
"inputs": {
"video": [
"139:247",
0
]
},
"_meta": {
"title": "Get Video Components"
}
},
"120": {
"class_type": "GetVideoComponents",
"inputs": {
"video": [
"141:250",
0
]
},
"_meta": {
"title": "Get Video Components"
}
},
"121": {
"class_type": "GetVideoComponents",
"inputs": {
"video": [
"144:253",
0
]
},
"_meta": {
"title": "Get Video Components"
}
},
"122": {
"class_type": "GetVideoComponents",
"inputs": {
"video": [
"146:256",
0
]
},
"_meta": {
"title": "Get Video Components"
}
},
"123": {
"class_type": "GetVideoComponents",
"inputs": {
"video": [
"148:259",
0
]
},
"_meta": {
"title": "Get Video Components"
}
},
"126": {
"class_type": "BatchImagesNode",
"inputs": {
"images.image0": [
"119",
0
],
"images.image1": [
"120",
0
],
"images.image2": [
"121",
0
],
"images.image3": [
"122",
0
],
"images.image4": [
"123",
0
]
},
"_meta": {
"title": "Batch Images"
}
},
"131": {
"class_type": "VHS_VideoCombine",
"inputs": {
"crf": 14,
"filename_prefix": "MJM/AIrt MAchIne/Final/MJM_AIrtMAchIne",
"format": "video/h264-mp4",
"frame_rate": [
"119",
2
],
"images": [
"126",
0
],
"loop_count": 0,
"pingpong": False,
"pix_fmt": "yuv420p",
"save_metadata": True,
"save_output": True,
"trim_to_audio": False
},
"_meta": {
"title": "Video Combine 🎥🅥🅗🅢"
}
},
"152": {
"class_type": "ImpactStringSelector",
"inputs": {
"multiline": False,
"select": 0,
"strings": [
"95",
0
]
},
"_meta": {
"title": "String Selector"
}
},
"153": {
"class_type": "ImpactStringSelector",
"inputs": {
"multiline": False,
"select": 1,
"strings": [
"95",
0
]
},
"_meta": {
"title": "String Selector"
}
},
"154": {
"class_type": "ImpactStringSelector",
"inputs": {
"multiline": False,
"select": 2,
"strings": [
"95",
0
]
},
"_meta": {
"title": "String Selector"
}
},
"155": {
"class_type": "ImpactStringSelector",
"inputs": {
"multiline": False,
"select": 3,
"strings": [
"95",
0
]
},
"_meta": {
"title": "String Selector"
}
},
"156": {
"class_type": "ImpactStringSelector",
"inputs": {
"multiline": False,
"select": 4,
"strings": [
"95",
0
]
},
"_meta": {
"title": "String Selector"
}
},
"161": {
"class_type": "SaveImage",
"inputs": {
"filename_prefix": "MJM/AIrt MAchIne/Frames/MJM_AIrtMAchIne_Im1",
"images": [
"5",
0
]
},
"_meta": {
"title": "Save Image"
}
},
"162": {
"class_type": "PrimitiveInt",
"inputs": {
"value": 1024
},
"_meta": {
"title": "Width"
}
},
"163": {
"class_type": "PrimitiveInt",
"inputs": {
"value": 1024
},
"_meta": {
"title": "Height"
}
},
"164": {
"class_type": "SaveImage",
"inputs": {
"filename_prefix": "MJM/AIrt MAchIne/Frames/MJM_AIrtMAchIne_Im2",
"images": [
"9",
0
]
},
"_meta": {
"title": "Save Image"
}
},
"165": {
"class_type": "SaveImage",
"inputs": {
"filename_prefix": "MJM/AIrt MAchIne/Frames/MJM_AIrtMAchIne_Im3",
"images": [
"14",
0
]
},
"_meta": {
"title": "Save Image"
}
},
"166": {
"class_type": "SaveImage",
"inputs": {
"filename_prefix": "MJM/AIrt MAchIne/Frames/MJM_AIrtMAchIne_Im4",
"images": [
"20",
0
]
},
"_meta": {
"title": "Save Image"
}
},
"167": {
"class_type": "SaveImage",
"inputs": {
"filename_prefix": "MJM/AIrt MAchIne/Frames/MJM_AIrtMAchIne_Im5",
"images": [
"25",
0
]
},
"_meta": {
"title": "Save Image"
}
},
"168": {
"class_type": "PreviewAny",
"inputs": {
"source": [
"95",
0
]
},
"_meta": {
"title": "Preview as Text"
}
},
"169": {
"class_type": "PreviewAny",
"inputs": {
"source": [
"152",
0
]
},
"_meta": {
"title": "Preview as Text"
}
},
"171": {
"class_type": "PreviewAny",
"inputs": {
"source": [
"153",
0
]
},
"_meta": {
"title": "Preview as Text"
}
},
"172": {
"class_type": "PreviewAny",
"inputs": {
"source": [
"154",
0
]
},
"_meta": {
"title": "Preview as Text"
}
},
"173": {
"class_type": "PreviewAny",
"inputs": {
"source": [
"155",
0
]
},
"_meta": {
"title": "Preview as Text"
}
},
"174": {
"class_type": "PreviewAny",
"inputs": {
"source": [
"156",
0
]
},
"_meta": {
"title": "Preview as Text"
}
},
"175": {
"class_type": "PreviewAny",
"inputs": {
"source": [
"139:132",
0
]
},
"_meta": {
"title": "Preview as Text"
}
},
"177": {
"class_type": "PreviewAny",
"inputs": {
"source": [
"141:249",
0
]
},
"_meta": {
"title": "Preview as Text"
}
},
"178": {
"class_type": "PreviewAny",
"inputs": {
"source": [
"144:252",
0
]
},
"_meta": {
"title": "Preview as Text"
}
},
"179": {
"class_type": "PreviewAny",
"inputs": {
"source": [
"146:255",
0
]
},
"_meta": {
"title": "Preview as Text"
}
},
"180": {
"class_type": "PreviewAny",
"inputs": {
"source": [
"148:258",
0
]
},
"_meta": {
"title": "Preview as Text"
}
},
"186": {
"class_type": "UNETLoader",
"inputs": {
"unet_name": "z_image_turbo_bf16.safetensors",
"weight_dtype": "default"
},
"_meta": {
"title": "Load Diffusion Model"
}
},
"187": {
"class_type": "CLIPLoader",
"inputs": {
"clip_name": "qwen_3_4b.safetensors",
"device": "default",
"type": "lumina2"
},
"_meta": {
"title": "Load CLIP"
}
},
"188": {
"class_type": "VAELoader",
"inputs": {
"vae_name": "ae.safetensors"
},
"_meta": {
"title": "Load VAE"
}
},
"103:196": {
"class_type": "KSampler",
"inputs": {
"cfg": 1,
"denoise": 1,
"latent_image": [
"103:197",
0
],
"model": [
"103:198",
0
],
"negative": [
"103:207",
0
],
"positive": [
"103:205",
0
],
"sampler_name": "res_multistep",
"scheduler": "simple",
"seed": 976857539749268,
"steps": 9
},
"_meta": {
"title": "KSampler"
}
},
"103:197": {
"class_type": "EmptySD3LatentImage",
"inputs": {
"batch_size": 1,
"height": [
"163",
0
],
"width": [
"162",
0
]
},
"_meta": {
"title": "EmptySD3LatentImage"
}
},
"103:198": {
"class_type": "ModelSamplingAuraFlow",
"inputs": {
"model": [
"186",
0
],
"sampling": "flow",
"shift": 3
},
"_meta": {
"title": "ModelSamplingAuraFlow"
}
},
"103:200": {
"class_type": "VAEDecode",
"inputs": {
"samples": [
"103:196",
0
],
"vae": [
"188",
0
]
},
"_meta": {
"title": "VAE Decode"
}
},
"103:201": {
"class_type": "VAEDecode",
"inputs": {
"samples": [
"103:202",
0
],
"vae": [
"188",
0
]
},
"_meta": {
"title": "VAE Decode"
}
},
"103:202": {
"class_type": "KSampler",
"inputs": {
"cfg": 1,
"denoise": 0.2,
"latent_image": [
"103:204",
0
],
"model": [
"103:198",
0
],
"negative": [
"103:207",
0
],
"positive": [
"103:205",
0
],
"sampler_name": "res_multistep",
"scheduler": "simple",
"seed": 974,
"steps": 9
},
"_meta": {
"title": "KSampler"
}
},
"103:203": {
"class_type": "UpscaleModelLoader",
"inputs": {
"model_name": "4x-UltraSharp.safetensors"
},
"_meta": {
"title": "Load Upscale Model"
}
},
"103:204": {
"class_type": "VAEEncode",
"inputs": {
"pixels": [
"103:98",
0
],
"vae": [
"188",
0
]
},
"_meta": {
"title": "VAE Encode"
}
},
"103:205": {
"class_type": "CLIPTextEncode",
"inputs": {
"clip": [
"187",
0
],
"text": [
"153",
0
]
},
"_meta": {
"title": "CLIP Text Encode (Prompt)"
}
},
"103:207": {
"class_type": "ConditioningZeroOut",
"inputs": {
"conditioning": [
"103:205",
0
]
},
"_meta": {
"title": "Conditioning Zero Out"
}
},
"103:98": {
"class_type": "ImageUpscaleWithModelBatched",
"inputs": {
"downscale_method": "bicubic",
"downscale_ratio": 0.5,
"images": [
"103:200",
0
],
"per_batch": 1,
"precision": "bfloat16",
"upscale_model": [
"103:203",
0
]
},
"_meta": {
"title": "Image Upscale With Model Batched"
}
},
"107:208": {
"class_type": "KSampler",
"inputs": {
"cfg": 1,
"denoise": 1,
"latent_image": [
"107:209",
0
],
"model": [
"107:210",
0
],
"negative": [
"107:220",
0
],
"positive": [
"107:218",
0
],
"sampler_name": "res_multistep",
"scheduler": "simple",
"seed": 689930766561209,
"steps": 9
},
"_meta": {
"title": "KSampler"
}
},
"107:209": {
"class_type": "EmptySD3LatentImage",
"inputs": {
"batch_size": 1,
"height": [
"163",
0
],
"width": [
"162",
0
]
},
"_meta": {
"title": "EmptySD3LatentImage"
}
},
"107:210": {
"class_type": "ModelSamplingAuraFlow",
"inputs": {
"model": [
"186",
0
],
"sampling": "flow",
"shift": 3
},
"_meta": {
"title": "ModelSamplingAuraFlow"
}
},
"107:212": {
"class_type": "ImageUpscaleWithModelBatched",
"inputs": {
"downscale_method": "bicubic",
"downscale_ratio": 0.5,
"images": [
"107:217",
0
],
"per_batch": 1,
"precision": "bfloat16",
"upscale_model": [
"107:216",
0
]
},
"_meta": {
"title": "Image Upscale With Model Batched"
}
},
"107:213": {
"class_type": "VAEDecode",
"inputs": {
"samples": [
"107:214",
0
],
"vae": [
"188",
0
]
},
"_meta": {
"title": "VAE Decode"
}
},
"107:214": {
"class_type": "KSampler",
"inputs": {
"cfg": 1,
"denoise": 0.2,
"latent_image": [
"107:215",
0
],
"model": [
"107:210",
0
],
"negative": [
"107:220",
0
],
"positive": [
"107:218",
0
],
"sampler_name": "res_multistep",
"scheduler": "simple",
"seed": 974,
"steps": 9
},
"_meta": {
"title": "KSampler"
}
},
"107:215": {
"class_type": "VAEEncode",
"inputs": {
"pixels": [
"107:212",
0
],
"vae": [
"188",
0
]
},
"_meta": {
"title": "VAE Encode"
}
},
"107:216": {
"class_type": "UpscaleModelLoader",
"inputs": {
"model_name": "4x-UltraSharp.safetensors"
},
"_meta": {
"title": "Load Upscale Model"
}
},
"107:217": {
"class_type": "VAEDecode",
"inputs": {
"samples": [
"107:208",
0
],
"vae": [
"188",
0
]
},
"_meta": {
"title": "VAE Decode"
}
},
"107:218": {
"class_type": "CLIPTextEncode",
"inputs": {
"clip": [
"187",
0
],
"text": [
"154",
0
]
},
"_meta": {
"title": "CLIP Text Encode (Prompt)"
}
},
"107:220": {
"class_type": "ConditioningZeroOut",
"inputs": {
"conditioning": [
"107:218",
0
]
},
"_meta": {
"title": "Conditioning Zero Out"
}
},
"111:221": {
"class_type": "KSampler",
"inputs": {
"cfg": 1,
"denoise": 1,
"latent_image": [
"111:222",
0
],
"model": [
"111:223",
0
],
"negative": [
"111:233",
0
],
"positive": [
"111:232",
0
],
"sampler_name": "res_multistep",
"scheduler": "simple",
"seed": 472700152594450,
"steps": 9
},
"_meta": {
"title": "KSampler"
}
},
"111:222": {
"class_type": "EmptySD3LatentImage",
"inputs": {
"batch_size": 1,
"height": [
"163",
0
],
"width": [
"162",
0
]
},
"_meta": {
"title": "EmptySD3LatentImage"
}
},
"111:223": {
"class_type": "ModelSamplingAuraFlow",
"inputs": {
"model": [
"186",
0
],
"sampling": "flow",
"shift": 3
},
"_meta": {
"title": "ModelSamplingAuraFlow"
}
},
"111:226": {
"class_type": "VAEDecode",
"inputs": {
"samples": [
"111:221",
0
],
"vae": [
"188",
0
]
},
"_meta": {
"title": "VAE Decode"
}
},
"111:227": {
"class_type": "VAEEncode",
"inputs": {
"pixels": [
"111:230",
0
],
"vae": [
"188",
0
]
},
"_meta": {
"title": "VAE Encode"
}
},
"111:228": {
"class_type": "KSampler",
"inputs": {
"cfg": 1,
"denoise": 0.2,
"latent_image": [
"111:227",
0
],
"model": [
"111:223",
0
],
"negative": [
"111:233",
0
],
"positive": [
"111:232",
0
],
"sampler_name": "res_multistep",
"scheduler": "simple",
"seed": 974,
"steps": 9
},
"_meta": {
"title": "KSampler"
}
},
"111:229": {
"class_type": "VAEDecode",
"inputs": {
"samples": [
"111:228",
0
],
"vae": [
"188",
0
]
},
"_meta": {
"title": "VAE Decode"
}
},
"111:230": {
"class_type": "ImageUpscaleWithModelBatched",
"inputs": {
"downscale_method": "bicubic",
"downscale_ratio": 0.5,
"images": [
"111:226",
0
],
"per_batch": 1,
"precision": "bfloat16",
"upscale_model": [
"111:231",
0
]
},
"_meta": {
"title": "Image Upscale With Model Batched"
}
},
"111:231": {
"class_type": "UpscaleModelLoader",
"inputs": {
"model_name": "4x-UltraSharp.safetensors"
},
"_meta": {
"title": "Load Upscale Model"
}
},
"111:232": {
"class_type": "CLIPTextEncode",
"inputs": {
"clip": [
"187",
0
],
"text": [
"155",
0
]
},
"_meta": {
"title": "CLIP Text Encode (Prompt)"
}
},
"111:233": {
"class_type": "ConditioningZeroOut",
"inputs": {
"conditioning": [
"111:232",
0
]
},
"_meta": {
"title": "Conditioning Zero Out"
}
},
"115:234": {
"class_type": "KSampler",
"inputs": {
"cfg": 1,
"denoise": 1,
"latent_image": [
"115:235",
0
],
"model": [
"115:236",
0
],
"negative": [
"115:246",
0
],
"positive": [
"115:245",
0
],
"sampler_name": "res_multistep",
"scheduler": "simple",
"seed": 579592923870881,
"steps": 9
},
"_meta": {
"title": "KSampler"
}
},
"115:235": {
"class_type": "EmptySD3LatentImage",
"inputs": {
"batch_size": 1,
"height": [
"163",
0
],
"width": [
"162",
0
]
},
"_meta": {
"title": "EmptySD3LatentImage"
}
},
"115:236": {
"class_type": "ModelSamplingAuraFlow",
"inputs": {
"model": [
"186",
0
],
"sampling": "flow",
"shift": 3
},
"_meta": {
"title": "ModelSamplingAuraFlow"
}
},
"115:239": {
"class_type": "VAEDecode",
"inputs": {
"samples": [
"115:234",
0
],
"vae": [
"188",
0
]
},
"_meta": {
"title": "VAE Decode"
}
},
"115:240": {
"class_type": "VAEDecode",
"inputs": {
"samples": [
"115:241",
0
],
"vae": [
"188",
0
]
},
"_meta": {
"title": "VAE Decode"
}
},
"115:241": {
"class_type": "KSampler",
"inputs": {
"cfg": 1,
"denoise": 0.2,
"latent_image": [
"115:242",
0
],
"model": [
"115:236",
0
],
"negative": [
"115:246",
0
],
"positive": [
"115:245",
0
],
"sampler_name": "res_multistep",
"scheduler": "simple",
"seed": 974,
"steps": 9
},
"_meta": {
"title": "KSampler"
}
},
"115:242": {
"class_type": "VAEEncode",
"inputs": {
"pixels": [
"115:243",
0
],
"vae": [
"188",
0
]
},
"_meta": {
"title": "VAE Encode"
}
},
"115:243": {
"class_type": "ImageUpscaleWithModelBatched",
"inputs": {
"downscale_method": "bicubic",
"downscale_ratio": 0.5,
"images": [
"115:239",
0
],
"per_batch": 1,
"precision": "bfloat16",
"upscale_model": [
"115:244",
0
]
},
"_meta": {
"title": "Image Upscale With Model Batched"
}
},
"115:244": {
"class_type": "UpscaleModelLoader",
"inputs": {
"model_name": "4x-UltraSharp.safetensors"
},
"_meta": {
"title": "Load Upscale Model"
}
},
"115:245": {
"class_type": "CLIPTextEncode",
"inputs": {
"clip": [
"187",
0
],
"text": [
"156",
0
]
},
"_meta": {
"title": "CLIP Text Encode (Prompt)"
}
},
"115:246": {
"class_type": "ConditioningZeroOut",
"inputs": {
"conditioning": [
"115:245",
0
]
},
"_meta": {
"title": "Conditioning Zero Out"
}
},
"139:132": {
"class_type": "GeminiNode",
"inputs": {
"images": [
"139:248",
0
],
"model": "gemini-3-pro-preview",
"prompt": "Write a wild prompt for a video model that will take first and last frame. Maximize motion and go crazy. Give me the prompt and nothing else.",
"seed": 590428414540815,
"system_prompt": ""
},
"_meta": {
"title": "Google Gemini"
}
},
"139:247": {
"class_type": "KlingFirstLastFrameNode",
"inputs": {
"duration": 3,
"end_frame": [
"9",
0
],
"first_frame": [
"5",
0
],
"generate_audio": False,
"model": "kling-v3",
"model.resolution": "1080p",
"prompt": [
"139:132",
0
],
"seed": 224701350
},
"_meta": {
"title": "Kling 3.0 First-Last-Frame to Video"
}
},
"139:248": {
"class_type": "ImageBatchMulti",
"inputs": {
"image_1": [
"5",
0
],
"image_2": [
"9",
0
],
"inputcount": 2
},
"_meta": {
"title": "Image Batch Multi"
}
},
"141:249": {
"class_type": "GeminiNode",
"inputs": {
"images": [
"141:251",
0
],
"model": "gemini-3-pro-preview",
"prompt": "Write a wild prompt for a video model that will take first and last frame. Maximize motion and go crazy. Give me the prompt and nothing else.",
"seed": 343805015166386,
"system_prompt": ""
},
"_meta": {
"title": "Google Gemini"
}
},
"141:250": {
"class_type": "KlingFirstLastFrameNode",
"inputs": {
"duration": 3,
"end_frame": [
"14",
0
],
"first_frame": [
"9",
0
],
"generate_audio": False,
"model": "kling-v3",
"model.resolution": "1080p",
"prompt": [
"141:249",
0
],
"seed": 277329452
},
"_meta": {
"title": "Kling 3.0 First-Last-Frame to Video"
}
},
"141:251": {
"class_type": "ImageBatchMulti",
"inputs": {
"image_1": [
"9",
0
],
"image_2": [
"14",
0
],
"inputcount": 2
},
"_meta": {
"title": "Image Batch Multi"
}
},
"144:252": {
"class_type": "GeminiNode",
"inputs": {
"images": [
"144:254",
0
],
"model": "gemini-3-pro-preview",
"prompt": "Write a wild prompt for a video model that will take first and last frame. Maximize motion and go crazy. Give me the prompt and nothing else.",
"seed": 152300153944343,
"system_prompt": ""
},
"_meta": {
"title": "Google Gemini"
}
},
"144:253": {
"class_type": "KlingFirstLastFrameNode",
"inputs": {
"duration": 3,
"end_frame": [
"20",
0
],
"first_frame": [
"14",
0
],
"generate_audio": False,
"model": "kling-v3",
"model.resolution": "1080p",
"prompt": [
"144:252",
0
],
"seed": 1510265499
},
"_meta": {
"title": "Kling 3.0 First-Last-Frame to Video"
}
},
"144:254": {
"class_type": "ImageBatchMulti",
"inputs": {
"image_1": [
"14",
0
],
"image_2": [
"20",
0
],
"inputcount": 2
},
"_meta": {
"title": "Image Batch Multi"
}
},
"146:255": {
"class_type": "GeminiNode",
"inputs": {
"images": [
"146:257",
0
],
"model": "gemini-3-pro-preview",
"prompt": "Write a wild prompt for a video model that will take first and last frame. Maximize motion and go crazy. Give me the prompt and nothing else.",
"seed": 27701164034917,
"system_prompt": ""
},
"_meta": {
"title": "Google Gemini"
}
},
"146:256": {
"class_type": "KlingFirstLastFrameNode",
"inputs": {
"duration": 3,
"end_frame": [
"25",
0
],
"first_frame": [
"20",
0
],
"generate_audio": False,
"model": "kling-v3",
"model.resolution": "1080p",
"prompt": [
"146:255",
0
],
"seed": 1843562002
},
"_meta": {
"title": "Kling 3.0 First-Last-Frame to Video"
}
},
"146:257": {
"class_type": "ImageBatchMulti",
"inputs": {
"image_1": [
"20",
0
],
"image_2": [
"25",
0
],
"inputcount": 2
},
"_meta": {
"title": "Image Batch Multi"
}
},
"148:140": {
"class_type": "ImageBatchMulti",
"inputs": {
"image_1": [
"25",
0
],
"image_2": [
"5",
0
],
"inputcount": 2
},
"_meta": {
"title": "Image Batch Multi"
}
},
"148:258": {
"class_type": "GeminiNode",
"inputs": {
"images": [
"148:140",
0
],
"model": "gemini-3-pro-preview",
"prompt": "Write a wild prompt for a video model that will take first and last frame. Maximize motion and go crazy. Give me the prompt and nothing else.",
"seed": 932797822206607,
"system_prompt": ""
},
"_meta": {
"title": "Google Gemini"
}
},
"148:259": {
"class_type": "KlingFirstLastFrameNode",
"inputs": {
"duration": 3,
"end_frame": [
"5",
0
],
"first_frame": [
"25",
0
],
"generate_audio": False,
"model": "kling-v3",
"model.resolution": "1080p",
"prompt": [
"148:258",
0
],
"seed": 1821078558
},
"_meta": {
"title": "Kling 3.0 First-Last-Frame to Video"
}
},
"99:101": {
"class_type": "UpscaleModelLoader",
"inputs": {
"model_name": "4x-UltraSharp.safetensors"
},
"_meta": {
"title": "Load Upscale Model"
}
},
"99:194": {
"class_type": "ImageUpscaleWithModelBatched",
"inputs": {
"downscale_method": "bicubic",
"downscale_ratio": 0.5,
"images": [
"99:73",
0
],
"per_batch": 1,
"precision": "bfloat16",
"upscale_model": [
"99:101",
0
]
},
"_meta": {
"title": "Image Upscale With Model Batched"
}
},
"99:195": {
"class_type": "KSampler",
"inputs": {
"cfg": 1,
"denoise": 0.2,
"latent_image": [
"99:97",
0
],
"model": [
"99:69",
0
],
"negative": [
"99:70",
0
],
"positive": [
"99:72",
0
],
"sampler_name": "res_multistep",
"scheduler": "simple",
"seed": 974,
"steps": 9
},
"_meta": {
"title": "KSampler"
}
},
"99:66": {
"class_type": "KSampler",
"inputs": {
"cfg": 1,
"denoise": 1,
"latent_image": [
"99:79",
0
],
"model": [
"99:69",
0
],
"negative": [
"99:70",
0
],
"positive": [
"99:72",
0
],
"sampler_name": "res_multistep",
"scheduler": "simple",
"seed": 251301300660602,
"steps": 9
},
"_meta": {
"title": "KSampler"
}
},
"99:69": {
"class_type": "ModelSamplingAuraFlow",
"inputs": {
"model": [
"186",
0
],
"sampling": "flow",
"shift": 3
},
"_meta": {
"title": "ModelSamplingAuraFlow"
}
},
"99:70": {
"class_type": "ConditioningZeroOut",
"inputs": {
"conditioning": [
"99:72",
0
]
},
"_meta": {
"title": "Conditioning Zero Out"
}
},
"99:72": {
"class_type": "CLIPTextEncode",
"inputs": {
"clip": [
"187",
0
],
"text": [
"152",
0
]
},
"_meta": {
"title": "CLIP Text Encode (Prompt)"
}
},
"99:73": {
"class_type": "VAEDecode",
"inputs": {
"samples": [
"99:66",
0
],
"vae": [
"188",
0
]
},
"_meta": {
"title": "VAE Decode"
}
},
"99:77": {
"class_type": "VAEDecode",
"inputs": {
"samples": [
"99:195",
0
],
"vae": [
"188",
0
]
},
"_meta": {
"title": "VAE Decode"
}
},
"99:79": {
"class_type": "EmptySD3LatentImage",
"inputs": {
"batch_size": 1,
"height": [
"163",
0
],
"width": [
"162",
0
]
},
"_meta": {
"title": "EmptySD3LatentImage"
}
},
"99:97": {
"class_type": "VAEEncode",
"inputs": {
"pixels": [
"99:194",
0
],
"vae": [
"188",
0
]
},
"_meta": {
"title": "VAE Encode"
}
}
})
# run() also passes api_key to the partner nodes in this workflow.
job = client.run(workflow, api_key=api_key)Questions & answers
High-resolution images with clear subjects tend to produce the best results, as they provide more detail for the AI models to work with.
Yes, the workflow includes a prompt generator that uses an LLM, which can be adjusted for more tailored video outputs.
Yes, you can bypass the save frames and save videos groups to output only the final combined video.
While the workflow can run locally, using a machine with a good GPU will enhance performance, especially when handling high-resolution images and video generation.