zsxkib/instant-id

Make realistic images of real people instantly

Capabilities

Reference ImagesNegative PromptSeed

Cost

Community model (estimated from hardware time)

Input Parameters

Name	Type	Description	Default	Constraints
`image`*	string(uri)	Input face image	`—`	—
`canny_strength`	number	Canny ControlNet strength, effective only if `enable_canny_controlnet` is true	`0.3`	min: 0, max: 1
`controlnet_conditioning_scale`	number	Scale for IdentityNet strength (for fidelity)	`0.8`	min: 0, max: 1.5
`depth_strength`	number	Depth ControlNet strength, effective only if `enable_depth_controlnet` is true	`0.5`	min: 0, max: 1
`disable_safety_checker`	boolean	Disable safety checker for generated images	`false`	—
`enable_canny_controlnet`	boolean	Enable Canny ControlNet, overrides strength if set to false	`false`	—
`enable_depth_controlnet`	boolean	Enable Depth ControlNet, overrides strength if set to false	`false`	—
`enable_lcm`	boolean	Enable Fast Inference with LCM (Latent Consistency Models) - speeds up inference steps, trade-off is the quality of the generated image. Performs better with close-up portrait face images	`false`	—
`enable_pose_controlnet`	boolean	Enable Openpose ControlNet, overrides strength if set to false	`true`	—
`enhance_nonface_region`	boolean	Enhance non-face region	`true`	—
`face_detection_input_height`	integer	Height of the input image for face detection	`640`	min: 640, max: 4096
`face_detection_input_width`	integer	Width of the input image for face detection	`640`	min: 640, max: 4096
`guidance_scale`	number	Scale for classifier-free guidance	`7.5`	min: 1, max: 50
`ip_adapter_scale`	number	Scale for image adapter strength (for detail)	`0.8`	min: 0, max: 1.5
`lcm_guidance_scale`	number	Only used when `enable_lcm` is set to True, Scale for classifier-free guidance when using LCM	`1.5`	min: 1, max: 20
`lcm_num_inference_steps`	integer	Only used when `enable_lcm` is set to True, Number of denoising steps when using LCM	`5`	min: 1, max: 10
`negative_prompt`	string	Input Negative Prompt	`""`	—
`num_inference_steps`	integer	Number of denoising steps	`30`	min: 1, max: 500
`num_outputs`	integer	Number of images to output	`1`	min: 1, max: 8
`output_format`	string	Format of the output images	`"webp"`	webpjpgpng
`output_quality`	integer	Quality of the output images, from 0 to 100. 100 is best quality, 0 is lowest quality.	`80`	min: 0, max: 100
`pose_image`	string(uri)	(Optional) reference pose image	`—`	—
`pose_strength`	number	Openpose ControlNet strength, effective only if `enable_pose_controlnet` is true	`0.4`	min: 0, max: 1
`prompt`	string	Input prompt	`"a person"`	—
`scheduler`	string	Scheduler	`"EulerDiscreteScheduler"`	DEISMultistepSchedulerHeunDiscreteSchedulerEulerDiscreteSchedulerDPMSolverMultistepSchedulerDPMSolverMultistepScheduler-KarrasDPMSolverMultistepScheduler-Karras-SDE
`sdxl_weights`	string	Pick which base weights you want to use	`"stable-diffusion-xl-base-1.0"`	stable-diffusion-xl-base-1.0juggernaut-xl-v8afrodite-xl-v2albedobase-xl-20albedobase-xl-v13animagine-xl-30anime-art-diffusion-xlanime-illust-diffusion-xldreamshaper-xldynavision-xl-v0610guofeng4-xlnightvision-xl-0791omnigen-xlpony-diffusion-v6-xlprotovision-xl-high-fidelRealVisXL_V3.0_TurboRealVisXL_V4.0_Lightning
`seed`	integer	Random seed. Leave blank to randomize the seed	`—`	—

imagerequiredstring

Input face image

canny_strengthnumber

Canny ControlNet strength, effective only if `enable_canny_controlnet` is true

Default: 0.3min: 0, max: 1

controlnet_conditioning_scalenumber

Scale for IdentityNet strength (for fidelity)

Default: 0.8min: 0, max: 1.5

depth_strengthnumber

Depth ControlNet strength, effective only if `enable_depth_controlnet` is true

Default: 0.5min: 0, max: 1

disable_safety_checkerboolean

Disable safety checker for generated images

Default: false

enable_canny_controlnetboolean

Enable Canny ControlNet, overrides strength if set to false

Default: false

enable_depth_controlnetboolean

Enable Depth ControlNet, overrides strength if set to false

Default: false

enable_lcmboolean

Enable Fast Inference with LCM (Latent Consistency Models) - speeds up inference steps, trade-off is the quality of the generated image. Performs better with close-up portrait face images

Default: false

enable_pose_controlnetboolean

Enable Openpose ControlNet, overrides strength if set to false

Default: true

enhance_nonface_regionboolean

Enhance non-face region

Default: true

face_detection_input_heightinteger

Height of the input image for face detection

Default: 640min: 640, max: 4096

face_detection_input_widthinteger

Width of the input image for face detection

Default: 640min: 640, max: 4096

guidance_scalenumber

Scale for classifier-free guidance

Default: 7.5min: 1, max: 50

ip_adapter_scalenumber

Scale for image adapter strength (for detail)

Default: 0.8min: 0, max: 1.5

lcm_guidance_scalenumber

Only used when `enable_lcm` is set to True, Scale for classifier-free guidance when using LCM

Default: 1.5min: 1, max: 20

lcm_num_inference_stepsinteger

Only used when `enable_lcm` is set to True, Number of denoising steps when using LCM

Default: 5min: 1, max: 10

negative_promptstring

Input Negative Prompt

Default: ""

num_inference_stepsinteger

Number of denoising steps

Default: 30min: 1, max: 500

num_outputsinteger

Number of images to output

Default: 1min: 1, max: 8

output_formatstring

Format of the output images

Default: "webp"

webpjpgpng

output_qualityinteger

Quality of the output images, from 0 to 100. 100 is best quality, 0 is lowest quality.

Default: 80min: 0, max: 100

pose_imagestring

(Optional) reference pose image

pose_strengthnumber

Openpose ControlNet strength, effective only if `enable_pose_controlnet` is true

Default: 0.4min: 0, max: 1

promptstring

Input prompt

Default: "a person"

schedulerstring

Scheduler

Default: "EulerDiscreteScheduler"

DEISMultistepSchedulerHeunDiscreteSchedulerEulerDiscreteSchedulerDPMSolverMultistepSchedulerDPMSolverMultistepScheduler-KarrasDPMSolverMultistepScheduler-Karras-SDE

sdxl_weightsstring

Pick which base weights you want to use

Default: "stable-diffusion-xl-base-1.0"

stable-diffusion-xl-base-1.0juggernaut-xl-v8afrodite-xl-v2albedobase-xl-20albedobase-xl-v13animagine-xl-30anime-art-diffusion-xlanime-illust-diffusion-xldreamshaper-xldynavision-xl-v0610guofeng4-xlnightvision-xl-0791omnigen-xlpony-diffusion-v6-xlprotovision-xl-high-fidelRealVisXL_V3.0_TurboRealVisXL_V4.0_Lightning

seedinteger

Random seed. Leave blank to randomize the seed

Version: 2e4785a4d80dUpdated: 7/25/20261.0M runs