Nodes/ComfyUI-QING/vLLM-I2I丨API
ComfyUI Node

vLLM-I2I丨API

A ComfyUI node in QING/API with 20 inputs and 4 outputs.

By sheengoa·Created 12 months ago·Updated 3 months ago· 16
vLLM-I2I丨API
  • image
  • mask
  • image
  • raw_response
  • status
  • debug_info
prompt
negative_prompt
platformvLLM
modeldefault
base_urlhttp://127.0.0.1:8091
i2i_endpointchat_completions
width1024
height1024
num_inference_steps50
guidance_scale4.00
top_p0.90
repetition_penalty1.10
seed0
n1
response_formatb64_json
timeout_sec120
api_key
advanced_options
CategoryQING/API

Inputs (20)

NameTypeDefaultDescription
promptSTRING编辑指令
negative_promptSTRING负面提示词(edits 模式常用)
imageIMAGE输入图像
platformCOMBOvLLM服务平台(用于动态拉取 /v1/models)
modelSTRINGdefault模型 id
base_urlSTRINGhttp://127.0.0.1:8091服务根地址
i2i_endpointCOMBOchat_completions含 mask 时自动改用 images_edits;与官方 I2I 示例一致请用 chat_completions
widthSTRING1024宽度(字符串整数)
heightSTRING1024高度(字符串整数)
num_inference_stepsINT501–250
guidance_scaleFLOAT4.000–30chat 模式写入 extra_body.guidance_scale;更多参数请使用 advanced_options
top_pFLOAT0.900–1
repetition_penaltyFLOAT1.100.5–2
seedINT00–2147483647
nINT11–4仅 images_edits 生效;n>1 时输出 IMAGE 批次
response_formatCOMBOb64_json仅 images_edits 生效
timeout_secINT1205–600
maskoptMASK可选遮罩;连接时强制走 /v1/images/edits
api_keyoptSTRING留空时使用环境变量或 QING 配置
advanced_optionsoptSTRING高级选项节点输出的 JSON(可通过 custom_json 兜底扩展)

Outputs (4)

NameTypeDescription
imageIMAGE
raw_responseSTRING
statusSTRING
debug_infoSTRING