Instant ID Generate Avatar
Generate personalized avatars from face photos with extensive ControlNet options (pose, canny, depth), multiple SDXL base models, and configurable identity preservation strength.
Quick Start
Requires an each::labs API key. Get one at eachlabs.ai.
curl -X POST https://api.eachlabs.ai/v1/prediction \
-H "Content-Type: application/json" \
-H "X-API-Key: $EACHLABS_API_KEY" \
-d '{
"model": "instant-id",
"version": "0.0.1",
"input": {
"image": "https://example.com/face-photo.jpg",
"prompt": "a person as a fantasy warrior in golden armor, epic background",
"num_inference_steps": 4,
"guidance_scale": 7.5,
"controlnet_conditioning_scale": 0.8,
"ip_adapter_scale": 0.8
}
}'
Parameters
| Parameter |
Type |
Default |
Description |
| canny_strength |
number |
0.3 |
Canny ControlNet strength (only if enable_canny_controlnet is true) |
| controlnet_conditioning_scale |
number |
0.8 |
Scale for IdentityNet strength (for fidelity) |
| depth_strength |
number |
0.5 |
Depth ControlNet strength (only if enable_depth_controlnet is true) |
| disable_safety_checker |
boolean |
false |
Disable safety checker |
| enable_canny_controlnet |
boolean |
false |
Enable canny edge detection control |
| enable_depth_controlnet |
boolean |
false |
Enable depth control |
| enable_lcm |
boolean |
false |
Enable LCM for faster inference |
| enable_pose_controlnet |
boolean |
true |
Enable Openpose ControlNet |
| enhance_nonface_region |
boolean |
true |
Enhance non-face region |
| guidance_scale |
number |
7.5 |
Scale for classifier-free guidance |
| height |
integer |
640 |
Height of output image |
| image |
string |
- |
Input face image |
| ip_adapter_scale |
number |
0.8 |
Scale for image adapter strength (for detail) |
| lcm_guidance_scale |
number |
1.5 |
Guidance scale when using LCM |
| lcm_num_inference_steps |
integer |
5 |
Steps when using LCM |
| negative_prompt |
string |
- |
Negative prompt |
| num_inference_steps |
integer |
4 |
Number of denoising steps |
| pose_image |
string |
- |
Optional reference pose image |
| pose_strength |
number |
0.4 |
Openpose ControlNet strength |
| prompt |
string |
a person |
Input prompt |
| scheduler |
string |
EulerDiscreteScheduler |
Scheduler. enum: DEISMultistepScheduler, HeunDiscreteScheduler, EulerDiscreteScheduler, DPMSolverMultistepScheduler, DPMSolverMultistepScheduler-Karras, DPMSolverMultistepScheduler-Karras-SDE |
| sdxl_weights |
string |
stable-diffusion-xl-base-1.0 |
Base model. enum: stable-diffusion-xl-base-1.0, juggernaut-xl-v8, afrodite-xl-v2, albedobase-xl-20, albedobase-xl-v13, animagine-xl-30, anime-art-diffusion-xl, anime-illust-diffusion-xl, dreamshaper-xl, dynavision-xl-v0610, guofeng4-xl, nightvision-xl-0791, omnigen-xl, pony-diffusion-v6-xl, protovision-xl-high-fidel |
| seed |
integer |
- |
Random seed |
| width |
integer |
640 |
Width of output image |
Examples
Anime avatar with custom base model:
curl -X POST https://api.eachlabs.ai/v1/prediction \
-H "Content-Type: application/json" \
-H "X-API-Key: $EACHLABS_API_KEY" \
-d '{
"model": "instant-id",
"version": "0.0.1",
"input": {
"image": "https://example.com/selfie.jpg",
"prompt": "a person as an anime character, colorful, detailed, vibrant background",
"sdxl_weights": "animagine-xl-30",
"controlnet_conditioning_scale": 0.7,
"ip_adapter_scale": 0.9,
"num_inference_steps": 6,
"guidance_scale": 8,
"width": 640,
"height": 640
}
}'
Pose-guided avatar with depth:
curl -X POST https://api.eachlabs.ai/v1/prediction \
-H "Content-Type: application/json" \
-H "X-API-Key: $EACHLABS_API_KEY" \
-d '{
"model": "instant-id",
"version": "0.0.1",
"input": {
"image": "https://example.com/face.jpg",
"pose_image": "https://example.com/pose-reference.jpg",
"prompt": "a person as a cyberpunk hacker in neon-lit room, detailed, high quality",
"enable_pose_controlnet": true,
"enable_depth_controlnet": true,
"pose_strength": 0.5,
"depth_strength": 0.4,
"guidance_scale": 7.5,
"num_inference_steps": 5,
"seed": 42
}
}'
Related Models
Documentation
1---2name: instant-id3description: Instant ID Generate Avatar | AI Avatar Generation. Generate personalized avatars from face photos with extensive controlnet and model options. Triggers: instant id, avatar generator, face avatar, id avatar, personalized avatar4---56# Instant ID Generate Avatar78Generate personalized avatars from face photos with extensive ControlNet options (pose, canny, depth), multiple SDXL base models, and configurable identity preservation strength.910## Quick Start1112> Requires an each::labs API key. Get one at [eachlabs.ai](https://eachlabs.ai).1314```bash15curl -X POST https://api.eachlabs.ai/v1/prediction \16 -H "Content-Type: application/json" \17 -H "X-API-Key: $EACHLABS_API_KEY" \18 -d '{19 "model": "instant-id",20 "version": "0.0.1",21 "input": {22 "image": "https://example.com/face-photo.jpg",23 "prompt": "a person as a fantasy warrior in golden armor, epic background",24 "num_inference_steps": 4,25 "guidance_scale": 7.5,26 "controlnet_conditioning_scale": 0.8,27 "ip_adapter_scale": 0.828 }29 }'30```3132## Parameters3334| Parameter | Type | Default | Description |35|-----------|------|---------|-------------|36| canny_strength | number | 0.3 | Canny ControlNet strength (only if enable_canny_controlnet is true) |37| controlnet_conditioning_scale | number | 0.8 | Scale for IdentityNet strength (for fidelity) |38| depth_strength | number | 0.5 | Depth ControlNet strength (only if enable_depth_controlnet is true) |39| disable_safety_checker | boolean | false | Disable safety checker |40| enable_canny_controlnet | boolean | false | Enable canny edge detection control |41| enable_depth_controlnet | boolean | false | Enable depth control |42| enable_lcm | boolean | false | Enable LCM for faster inference |43| enable_pose_controlnet | boolean | true | Enable Openpose ControlNet |44| enhance_nonface_region | boolean | true | Enhance non-face region |45| guidance_scale | number | 7.5 | Scale for classifier-free guidance |46| height | integer | 640 | Height of output image |47| image | string | - | Input face image |48| ip_adapter_scale | number | 0.8 | Scale for image adapter strength (for detail) |49| lcm_guidance_scale | number | 1.5 | Guidance scale when using LCM |50| lcm_num_inference_steps | integer | 5 | Steps when using LCM |51| negative_prompt | string | - | Negative prompt |52| num_inference_steps | integer | 4 | Number of denoising steps |53| pose_image | string | - | Optional reference pose image |54| pose_strength | number | 0.4 | Openpose ControlNet strength |55| prompt | string | a person | Input prompt |56| scheduler | string | EulerDiscreteScheduler | Scheduler. enum: DEISMultistepScheduler, HeunDiscreteScheduler, EulerDiscreteScheduler, DPMSolverMultistepScheduler, DPMSolverMultistepScheduler-Karras, DPMSolverMultistepScheduler-Karras-SDE |57| sdxl_weights | string | stable-diffusion-xl-base-1.0 | Base model. enum: stable-diffusion-xl-base-1.0, juggernaut-xl-v8, afrodite-xl-v2, albedobase-xl-20, albedobase-xl-v13, animagine-xl-30, anime-art-diffusion-xl, anime-illust-diffusion-xl, dreamshaper-xl, dynavision-xl-v0610, guofeng4-xl, nightvision-xl-0791, omnigen-xl, pony-diffusion-v6-xl, protovision-xl-high-fidel |58| seed | integer | - | Random seed |59| width | integer | 640 | Width of output image |6061## Examples6263**Anime avatar with custom base model:**64```bash65curl -X POST https://api.eachlabs.ai/v1/prediction \66 -H "Content-Type: application/json" \67 -H "X-API-Key: $EACHLABS_API_KEY" \68 -d '{69 "model": "instant-id",70 "version": "0.0.1",71 "input": {72 "image": "https://example.com/selfie.jpg",73 "prompt": "a person as an anime character, colorful, detailed, vibrant background",74 "sdxl_weights": "animagine-xl-30",75 "controlnet_conditioning_scale": 0.7,76 "ip_adapter_scale": 0.9,77 "num_inference_steps": 6,78 "guidance_scale": 8,79 "width": 640,80 "height": 64081 }82 }'83```8485**Pose-guided avatar with depth:**86```bash87curl -X POST https://api.eachlabs.ai/v1/prediction \88 -H "Content-Type: application/json" \89 -H "X-API-Key: $EACHLABS_API_KEY" \90 -d '{91 "model": "instant-id",92 "version": "0.0.1",93 "input": {94 "image": "https://example.com/face.jpg",95 "pose_image": "https://example.com/pose-reference.jpg",96 "prompt": "a person as a cyberpunk hacker in neon-lit room, detailed, high quality",97 "enable_pose_controlnet": true,98 "enable_depth_controlnet": true,99 "pose_strength": 0.5,100 "depth_strength": 0.4,101 "guidance_scale": 7.5,102 "num_inference_steps": 5,103 "seed": 42104 }105 }'106```107108## Related Models109110- [instant-id-ip-adapter](../instant-id-ip-adapter/) - Styled avatar generation (3D, Emoji, Pixels)111- [photomaker](../photomaker/) - Identity-preserving photo generation112- [become-image](../become-image/) - Style transfer with identity preservation113114## Documentation115116- [each::labs Docs](https://docs.eachlabs.ai)117- [API Reference](https://docs.eachlabs.ai/api/overview)
awesome-genmedia