executing-plans
executing-plans
Use when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available
each-upscalerImage Upscaler by Each AI | AI Image Upscaling. Upscale images with creativity and resemblance controls, face enhancement, and multiple SD models. Triggers: upscaler, upscale, image upscale, each upscaler, enhance image, super resolution
Independent GitHub snapshot · Automated classification · Original documentation governs
Upscale images with advanced controls for creativity, resemblance, sharpening, and optional face/hand enhancement. Supports multiple Stable Diffusion base models and schedulers.
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": "each-upscaler",
"version": "0.0.1",
"input": {
"image": "https://example.com/low-res-photo.jpg",
"scale_factor": 2,
"creativity": 0.35,
"resemblance": 0.6,
"num_inference_steps": 18
}
}'
| Parameter | Type | Default | Description |
|---|---|---|---|
| creativity | number | 0.35 | Creativity level, try from 0.3 - 0.9 |
| custom_sd_model | string | - | Custom SD model URL |
| downscaling | boolean | false | Downscale before upscaling. Can improve quality for large images |
| downscaling_resolution | integer | 768 | Downscaling resolution |
| dynamic | number | 6 | HDR effect, try from 3 - 9 |
| handfix | string | disabled | Hand fixing mode. enum: disabled, hands_only, image_and_hands |
| image | string | - | Input image |
| lora_links | string | - | Link to LoRA file(s) for upscaling |
| mask | string | - | Mask image to mark areas to preserve during upscaling |
| negative_prompt | string | (worst quality, low quality, normal quality:2)... | Negative prompt |
| num_inference_steps | integer | 18 | Number of denoising steps |
| output_format | string | png | Output format. enum: webp, jpg, png |
| pattern | boolean | false | Upscale a pattern with seamless tiling |
| prompt | string | masterpiece, best quality, highres... | Prompt |
| resemblance | number | 0.6 | Resemblance to original, try from 0.3 - 1.6 |
| scale_factor | number | 2 | Scale factor |
| scheduler | string | DPM++ 3M SDE Karras | Scheduler. enum: DPM++ 2M Karras, DPM++ SDE Karras, DPM++ 2M SDE Exponential, DPM++ 2M SDE Karras, Euler a, Euler, LMS, Heun, DPM2, DPM2 a, DPM++ 2S a, DPM++ 2M, DPM++ SDE, DPM++ 2M SDE, DPM++ 2M SDE Heun, and more |
| sd_model | string | juggernaut_reborn.safetensors | Base model. enum: epicrealism_naturalSinRC1VAE.safetensors, juggernaut_reborn.safetensors, flat2DAnimerge_v45Sharp.safetensors |
| seed | integer | 1337 | Random seed |
| sharpen | number | 0 | Sharpening intensity after upscaling |
| tiling_height | integer | 144 | Tiling height |
| tiling_width | integer | 112 | Tiling width |
High-quality portrait upscale with face fix:
curl -X POST https://api.eachlabs.ai/v1/prediction \
-H "Content-Type: application/json" \
-H "X-API-Key: $EACHLABS_API_KEY" \
-d '{
"model": "each-upscaler",
"version": "0.0.1",
"input": {
"image": "https://example.com/portrait.jpg",
"scale_factor": 2,
"creativity": 0.4,
"resemblance": 0.8,
"handfix": "image_and_hands",
"dynamic": 5,
"output_format": "png"
}
}'
Creative upscale with HDR:
curl -X POST https://api.eachlabs.ai/v1/prediction \
-H "Content-Type: application/json" \
-H "X-API-Key: $EACHLABS_API_KEY" \
-d '{
"model": "each-upscaler",
"version": "0.0.1",
"input": {
"image": "https://example.com/landscape.jpg",
"scale_factor": 4,
"creativity": 0.6,
"resemblance": 0.5,
"dynamic": 8,
"sharpen": 1.5,
"sd_model": "epicrealism_naturalSinRC1VAE.safetensors [84d76a0328]",
"num_inference_steps": 25
}
}'