MOSS-Image
Generate images with MOSS-Image, and compare every CFG, steps, sampler and scheduler on one prompt.
Télécharger le workflowEssayer dans Cloud Documentation de l’API
Loading the grid…
Explore more grids
Your output will appear here.
Dans le workflow
Un aperçu en lecture seule du modèle source. Ouvrez-le dans Cloud pour explorer les nœuds et apporter des modifications.
Loading the graph
Open full-size workflow preview- Comfy Cloud
- Oui
- Sortie
- Image
- Auteur
- Auteur indisponible
Appelez ce modèle depuis votre code
La requête ci-dessous utilise le même graphe préparé et les mêmes paramètres que le Playground. Chaque envoi peut lancer une exécution payante ; ne relancez pas automatiquement un envoi dont l’issue est inconnue.
1 Obtenir une clé API ↗
What you need
- Runs on
- Comfy Cloud
- Clé
- COMFY_API_KEY
Nécessite un forfait Cloud payant et des crédits.
Télécharger le JSON de l’APIDocumentation de l’API ↗
# 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({
"1": {
"class_type": "UNETLoader",
"inputs": {
"unet_name": "moss_image_sft_int8_convrot.safetensors",
"weight_dtype": "default"
},
"_meta": {
"title": "Load Diffusion Model"
}
},
"2": {
"class_type": "CLIPLoader",
"inputs": {
"clip_name": "qwen3.5_9b_w4a8_gptq.safetensors",
"device": "default",
"type": "moss_image"
},
"_meta": {
"title": "Load CLIP"
}
},
"3": {
"class_type": "VAELoader",
"inputs": {
"vae_name": "flux2-vae_small.safetensors"
},
"_meta": {
"title": "Load VAE"
}
},
"4": {
"class_type": "CLIPTextEncode",
"inputs": {
"clip": [
"2",
0
],
"text": "Epic wide landscape photograph of a sprawling tropical karst landscape in spring rain: hundreds of steep limestone towers covered in lush green vegetation rising out of morning mist, a wide jade river winding between flooded rice terraces, soft diffused light breaking through low clouds, immense scale and depth, ultra-detailed, large format landscape photography, no people, no text."
},
"_meta": {
"title": "Positive Prompt"
}
},
"5": {
"class_type": "CLIPTextEncode",
"inputs": {
"clip": [
"2",
0
],
"text": "nsfw, worst quality, bad anatomy, watermark, jpeg artifact"
},
"_meta": {
"title": "Negative Prompt"
}
},
"6": {
"class_type": "EmptyFlux2LatentImage",
"inputs": {
"batch_size": 1,
"height": 768,
"width": 1344
},
"_meta": {
"title": "Empty Flux 2 Latent"
}
},
"90": {
"class_type": "PrimitiveFloat",
"inputs": {
"value": 3
},
"_meta": {
"title": "CFG (sweep)"
}
},
"91": {
"class_type": "PrimitiveInt",
"inputs": {
"value": 25
},
"_meta": {
"title": "Steps (sweep)"
}
},
"92": {
"class_type": "KSamplerSelect",
"inputs": {
"sampler_name": "euler"
},
"_meta": {
"title": "Sampler (sweep)"
}
},
"93": {
"class_type": "CFGGuider",
"inputs": {
"cfg": [
"90",
0
],
"model": [
"1",
0
],
"negative": [
"5",
0
],
"positive": [
"4",
0
]
},
"_meta": {
"title": "CFG Guider"
}
},
"94": {
"class_type": "RandomNoise",
"inputs": {
"noise_seed": 42
},
"_meta": {
"title": "Noise (seed 42)"
}
},
"100": {
"class_type": "BasicScheduler",
"inputs": {
"denoise": 1,
"model": [
"1",
0
],
"scheduler": "simple",
"steps": [
"91",
0
]
},
"_meta": {
"title": "Scheduler simple"
}
},
"200": {
"class_type": "SamplerCustomAdvanced",
"inputs": {
"guider": [
"93",
0
],
"latent_image": [
"6",
0
],
"noise": [
"94",
0
],
"sampler": [
"92",
0
],
"sigmas": [
"100",
0
]
},
"_meta": {
"title": "Sample simple"
}
},
"400": {
"class_type": "VAEDecode",
"inputs": {
"samples": [
"200",
0
],
"vae": [
"3",
0
]
},
"_meta": {
"title": "Decode simple"
}
},
"599": {
"class_type": "SaveImage",
"inputs": {
"filename_prefix": "moss_strip",
"images": [
"400",
0
]
},
"_meta": {
"title": "Save Image"
}
}
})
# 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("599"):
output.to_file(output.name)