Flux API , payload for the text2img API

[ngochoang0105]

How can I add all the "VAE / Text Encoder" parameters to the API payload? I can't find it in the documentation! The result is always a black image. Here is my code.
`import requests
import base64
from PIL import Image
from io import BytesIO
import os
import json

web_ui_url = "http://127.0.0.1:7860"

payload = {
"prompt": "1girl",
"seed": -1,
"steps": 20,
"cfg_scale": 1,
"width": 512 ,
"height": 512,
"batch_size": 1,
"n_iter": 1,
"sampler_index": "Euler",
"schedule_type": "Simple",
"send_images": True,
"save_images": False,

"override_settings": {
    "sd_model_checkpoint": "flux1-dev.safetensors",
}

}

response = requests.post(f"{web_ui_url}/sdapi/v1/txt2img", json=payload)

if response.status_code == 200:
json_data = response.json()
image_base64 = json_data['images'][0]
image_bytes = base64.b64decode(image_base64)
image = Image.open(BytesIO(image_bytes))
image.show()`

[acleitao-projects]

did you end up figuring this out?

[acleitao-projects]

found it ... call the options end point with this payload sdapi/v1/options (in your case for flux in my case for sdxl)

{
  "sd_model_checkpoint": "juggernautXL_juggXIByRundiffusion.safetensors",
  "sd_vae": "sdxlVAE_sdxlVAE.safetensors",
  "forge_additional_modules": [
      "/opt/stable-diffusion-webui-forge/models/VAE/clip_l.safetensors",
      "/opt/stable-diffusion-webui-forge/models/VAE/t5xxl_fp16.safetensors"
    ]
}

Now looking how to unload the model immediately after finishing generation