Z Image | Turbo | Controlnet | Lora
Generate images with ControlNet-guided generation and LoRA weight support. Use canny edges, depth maps, or pose detection from reference images combined with custom LoRA models for precise control over composition and style.
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": "z-image-turbo-controlnet-lora",
"version": "0.0.1",
"input": {
"prompt": "A futuristic robot in the same pose, metallic chrome finish",
"image_url": "https://example.com/person-pose.jpg",
"preprocess": "pose",
"loras": [{"path": "https://example.com/scifi-lora.safetensors", "scale": 0.8}],
"control_scale": 0.9,
"image_size": "auto",
"output_format": "png"
}
}'
Parameters
| Parameter |
Type |
Default |
Description |
| acceleration |
string |
none |
The acceleration level to use. Options: none, regular, high |
| control_end |
number |
0.4 |
The end of the controlnet conditioning. |
| control_scale |
number |
0.9 |
The scale of the controlnet conditioning. |
| control_start |
number |
0 |
The start of the controlnet conditioning. |
| enable_prompt_expansion |
boolean |
false |
Whether to enable prompt expansion. |
| enable_safety_checker |
boolean |
true |
If set to true, the safety checker will be enabled. |
| image_size |
string |
auto |
The size of the generated image. Options: square_hd, square, portrait_4_3, portrait_16_9, landscape_4_3, landscape_16_9, auto |
| image_url |
string |
- |
URL of Image for ControlNet generation. |
| loras |
array |
- |
List of LoRA weights to apply. |
| num_images |
integer |
1 |
The number of images to generate. |
| num_inference_steps |
integer |
8 |
The number of inference steps to perform. |
| output_format |
string |
png |
The format of the generated image. Options: jpeg, png, webp |
| preprocess |
string |
none |
Preprocessing to apply to the image. Options: none, canny, depth, pose |
| prompt |
string |
- |
The prompt to generate an image from. |
| seed |
integer |
- |
Seed for reproducible generation. |
Examples
Canny edge-guided generation:
curl -X POST https://api.eachlabs.ai/v1/prediction \
-H "Content-Type: application/json" \
-H "X-API-Key: $EACHLABS_API_KEY" \
-d '{
"model": "z-image-turbo-controlnet-lora",
"version": "0.0.1",
"input": {
"prompt": "Elegant architectural rendering, glass and steel",
"image_url": "https://example.com/building-sketch.jpg",
"preprocess": "canny",
"loras": [{"path": "https://example.com/architecture-lora.safetensors", "scale": 0.7}]
}
}'
Depth-guided with fine-tuned control:
curl -X POST https://api.eachlabs.ai/v1/prediction \
-H "Content-Type: application/json" \
-H "X-API-Key: $EACHLABS_API_KEY" \
-d '{
"model": "z-image-turbo-controlnet-lora",
"version": "0.0.1",
"input": {
"prompt": "An enchanted forest scene with magical creatures",
"image_url": "https://example.com/forest-photo.jpg",
"preprocess": "depth",
"loras": [{"path": "https://example.com/fantasy-lora.safetensors", "scale": 0.9}],
"control_scale": 0.7,
"control_start": 0,
"control_end": 0.6,
"acceleration": "regular",
"num_images": 2,
"seed": 42
}
}'
Related Models
Documentation
1---2name: z-image-turbo-controlnet-lora3description: Z Image | Turbo | Controlnet | Lora. Generate images with ControlNet guidance and LoRA weights. Triggers: controlnet, lora, guided generation, z image, pose to image, depth4---56# Z Image | Turbo | Controlnet | Lora78Generate images with ControlNet-guided generation and LoRA weight support. Use canny edges, depth maps, or pose detection from reference images combined with custom LoRA models for precise control over composition and style.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": "z-image-turbo-controlnet-lora",20 "version": "0.0.1",21 "input": {22 "prompt": "A futuristic robot in the same pose, metallic chrome finish",23 "image_url": "https://example.com/person-pose.jpg",24 "preprocess": "pose",25 "loras": [{"path": "https://example.com/scifi-lora.safetensors", "scale": 0.8}],26 "control_scale": 0.9,27 "image_size": "auto",28 "output_format": "png"29 }30 }'31```3233## Parameters3435| Parameter | Type | Default | Description |36|-----------|------|---------|-------------|37| acceleration | string | none | The acceleration level to use. Options: `none`, `regular`, `high` |38| control_end | number | 0.4 | The end of the controlnet conditioning. |39| control_scale | number | 0.9 | The scale of the controlnet conditioning. |40| control_start | number | 0 | The start of the controlnet conditioning. |41| enable_prompt_expansion | boolean | false | Whether to enable prompt expansion. |42| enable_safety_checker | boolean | true | If set to true, the safety checker will be enabled. |43| image_size | string | auto | The size of the generated image. Options: `square_hd`, `square`, `portrait_4_3`, `portrait_16_9`, `landscape_4_3`, `landscape_16_9`, `auto` |44| image_url | string | - | URL of Image for ControlNet generation. |45| loras | array | - | List of LoRA weights to apply. |46| num_images | integer | 1 | The number of images to generate. |47| num_inference_steps | integer | 8 | The number of inference steps to perform. |48| output_format | string | png | The format of the generated image. Options: `jpeg`, `png`, `webp` |49| preprocess | string | none | Preprocessing to apply to the image. Options: `none`, `canny`, `depth`, `pose` |50| prompt | string | - | The prompt to generate an image from. |51| seed | integer | - | Seed for reproducible generation. |5253## Examples5455**Canny edge-guided generation:**56```bash57curl -X POST https://api.eachlabs.ai/v1/prediction \58 -H "Content-Type: application/json" \59 -H "X-API-Key: $EACHLABS_API_KEY" \60 -d '{61 "model": "z-image-turbo-controlnet-lora",62 "version": "0.0.1",63 "input": {64 "prompt": "Elegant architectural rendering, glass and steel",65 "image_url": "https://example.com/building-sketch.jpg",66 "preprocess": "canny",67 "loras": [{"path": "https://example.com/architecture-lora.safetensors", "scale": 0.7}]68 }69 }'70```7172**Depth-guided with fine-tuned control:**73```bash74curl -X POST https://api.eachlabs.ai/v1/prediction \75 -H "Content-Type: application/json" \76 -H "X-API-Key: $EACHLABS_API_KEY" \77 -d '{78 "model": "z-image-turbo-controlnet-lora",79 "version": "0.0.1",80 "input": {81 "prompt": "An enchanted forest scene with magical creatures",82 "image_url": "https://example.com/forest-photo.jpg",83 "preprocess": "depth",84 "loras": [{"path": "https://example.com/fantasy-lora.safetensors", "scale": 0.9}],85 "control_scale": 0.7,86 "control_start": 0,87 "control_end": 0.6,88 "acceleration": "regular",89 "num_images": 2,90 "seed": 4291 }92 }'93```9495## Related Models9697- [Z Image | Turbo | Lora](../z-image-turbo-lora/) - Text-to-image with LoRA98- [Z Image | Turbo | Image to Image | Lora](../z-image-turbo-image-to-image-lora/) - Image-to-image with LoRA99- [Z Image | Turbo | Controlnet](../z-image-turbo-controlnet/) - ControlNet without LoRA100101## Documentation102103- [each::labs Docs](https://docs.eachlabs.ai)104- [API Reference](https://docs.eachlabs.ai/api/overview)
awesome-genmedia