Skip to main content

Inpainting Endpoint

Overview​

Inpainting API is used to change (inpaint) some part of an image according to specific requirements, based on trained or on public models. Pass the appropriate request parameters to the endpoint.

This endpoint generates and returns an image from an image and a mask passed with their URLs in the request together with a model's ID.

You can also add your description of the desired result by passing prompt and negative prompt.

Inpainting endpoint result

Request​

--request POST 'https://modelslab.com/api/v6/images/inpaint' \

Make a POST request to https://modelslab.com/api/v6/images/inpaint endpoint and pass the required parameters as a request body.

Watch the Inpaint Community Model Endpoint Demo video to see it in action Postman.​

Attributes​

ParameterDescriptionValues
keyYour API Key used for request authorization.string
model_idThe ID of the model to be used. It can be a public model or one you have trained.id
promptA text description of what you want in the generated image.string
negative_promptItems you don't want in the image.string
init_imageLink to the initial image to be used as a reference.url
mask_imageLink to the mask image for inpainting.url
widthThe width of the image. Maximum value is 1024 pixels.integer
heightThe height of the image. Maximum value is 1024 pixels.integer
samplesThe number of images to be returned in the response. The maximum value is 4.integer
stepsThe number of denoising steps. Accepts values 21, 31, or 41.integer
safety_checkerA checker for NSFW images. If detected, such images will be replaced by a blank image. Default is "yes"."yes"/"no"
safety_checker_typeHow to modify the image if NSFW content is found. Default is "sensitive_content_text". Options: "blur", "sensitive_content_text", "pixelate", "black"."blur", "sensitive_content_text", "pixelate", "black"
enhance_promptEnhance prompts for better results. Default is "yes". Options: "yes"/"no"."yes"/"no"
enhance_styleAllows you to set the style of the image for better results. Available styles include:enhance, cinematic-diva, nude, nsfw, sex, abstract-expressionism, academia, action-figure, adorable-3d-character, adorable-kawaii, art-deco, art-nouveau, astral-aura, avant-garde, baroque, bauhaus-style-poster, blueprint-schematic-drawing, caricature, cel-shaded-art, character-design-sheet, classicism-art, color-field-painting, colored-pencil-art, conceptual-art, constructivism, cubism, dadaism, dark-fantasy, dark-moody-atmosphere, dmt-art, doodle-art, double-exposure, dripping-paint-splatter, expressionism, faded-polaroid-photo, fauvism, flat-2d, fortnite-style, futurism, glitchcore, glo-fi, googie-style, graffiti-art, harlem-renaissance-art, high-fashion, idyllic, impressionism, infographic-drawing, ink-dripping-drawing, japanese-ink-drawing, knolling-photography, light-cheery-atmosphere, logo-design, luxurious-elegance, macro-photography, mandola-art, marker-drawing, medievalism, minimalism, neo-baroque, neo-byzantine, neo-futurism, neo-impressionism, neo-rococo, neoclassicism, op-art, ornate-and-intricate, pencil-sketch-drawing, pop-art-2, rococo, silhouette-art, simple-vector-art, sketchup, steampunk-2, surrealism, suprematism, terragen, tranquil-relaxing-atmosphere, sticker-designs, vibrant-rim-light, volumetric-lighting, watercolor, whimsical-and-playful, sharp, masterpiece, photograph, negative, cinematic, ads-advertising, ads-automotive, ads-corporate, ads-fashion-editorial, ads-food-photography, ads-gourmet-food-photography, ads-luxury, ads-real-estate, ads-retail, abstract, abstract-expressionism, art-deco, art-nouveau, constructivist, cubist, expressionist, graffiti, hyperrealism, impressionist, pointillism, pop-art, psychedelic, renaissance, steampunk, surrealist, typography, watercolor, futuristic-biomechanical, futuristic-biomechanical-cyberpunk, futuristic-cybernetic, futuristic-cybernetic-robot, futuristic-cyberpunk-cityscape, futuristic-futuristic, futuristic-retro-cyberpunk, futuristic-retro, futuristic-sci-fi, futuristic-vaporwave, game-bubble, game-cyberpunk, game-fighting, game-gta, game-mario, game-minecraft, game-pokemon, game-retro-arcade, game-retro, game-rpg-fantasy, game-strategy, game-streetfighter, game-zelda, misc-architectural, misc-disco, misc-dreamscape, misc-dystopian, misc-fairy-tale, misc-gothic, misc-grunge, misc-horror, misc-kawaii, misc-lovecraftian, misc-macabre, misc-manga, misc-metropolis, misc-minimalist, misc-monochrome, misc-nautical, misc-space, misc-stained-glass, misc-techwear-fashion, misc-tribal, misc-zentangle, papercraft-collage, papercraft-flat-papercut, papercraft-kirigami, papercraft-paper-mache, papercraft-paper-quilling, papercraft-papercut-collage, papercraft-papercut-shadow-box, papercraft-stacked-papercut, papercraft-thick-layered-papercut, photo-alien, photo-film-noir, photo-glamour, photo-hdr, photo-iphone-photographic, photo-long-exposure, photo-neon-noir, photo-silhouette, photo-tilt-shift, 3d-model, analog-film, anime, cinematic, comic-book, craft-clay, digital-art, fantasy-art, isometric, line-art, lowpoly, neonpunk, origami, photographic, pixel-art, texture
guidance_scaleScale for classifier-free guidance. Minimum 1, Maximum 20.integer
tomesdEnable tomesd to generate images quickly. Default is "yes". Options: "yes"/"no"."yes"/"no"
use_karras_sigmasUse Keras sigmas to generate images. Produces nice results. Default is "yes". Options: "yes"/"no"."yes"/"no"
algorithm_typeUsed in the DPMSolverMultistepScheduler scheduler. Default is "none". Options: "dpmsolver+++"."dpmsolver+++"
vaeUse a custom VAE for generating images. Default is null.null
lora_strengthSpecify the strength of the LoRa model you're using. Range from 0.1 to 1, or comma-separated values.string
lora_modelPass LoRa model ID. Multiple LoRa models supported; comma-separated values. Example: "contrast-fix, yae-miko-genshin".string
strengthPrompt strength when using the initial image. Range from 0 to 1.float
ip_adapter_idIP adapter ID. Supported IDs are "ip-adapter_sdxl", "ip-adapter_sd15", "ip-adapter-plus-face_sd15".string
ip_adapter_scaleScale for the IP adapter. Should be between 0 and 1.float
ip_adapter_imageValid image URL for the IP adapter.url
schedulerUse this to set a scheduler.scheduler
seedUsed to reproduce results. The same seed will generate the same image again. Pass null for a random number.integer
webhookSet a URL to receive a POST API call once image generation is complete.url
track_idThis ID is returned in the response to the webhook API call and is used to identify the request.id
clip_skipClip Skip. Minimum 1, Maximum 8.integer
base64Get the response as a base64 string. Pass init_image and mask_image as base64 strings to get a base64 response. Default is "no"."yes"/"no"
tempCreate a temporary image link valid for 24 hours."yes"/"no"

Schedulers​

This endpoint also supports schedulers. Use the "scheduler" parameter in the request body to pass a specific scheduler from the list below:

  • DDPMScheduler
  • DDIMScheduler
  • PNDMScheduler
  • LMSDiscreteScheduler
  • EulerDiscreteScheduler
  • EulerAncestralDiscreteScheduler
  • DPMSolverMultistepScheduler
  • HeunDiscreteScheduler
  • KDPM2DiscreteScheduler
  • DPMSolverSinglestepScheduler
  • KDPM2AncestralDiscreteScheduler
  • UniPCMultistepScheduler
  • DDIMInverseScheduler
  • DEISMultistepScheduler
  • IPNDMScheduler
  • KarrasVeScheduler
  • ScoreSdeVeScheduler
  • LCMScheduler

Example​

Body​

Body
{
  "key": "",
  "model_id": "your_model_id",
  "prompt": "a cat sitting on a bench",
  "negative_prompt": null,
  "init_image": "https://raw.githubusercontent.com/CompVis/stable-diffusion/main/data/inpainting_examples/overture-creations-5sI6fQgYIuo.png",
  "mask_image": "https://raw.githubusercontent.com/CompVis/stable-diffusion/main/data/inpainting_examples/overture-creations-5sI6fQgYIuo_mask.png",
  "width": "512",
  "height": "512",
  "samples": "1",
  "steps": "21",
  "safety_checker": "no",
  "enhance_prompt": "yes",
  "guidance_scale": 7.5,
  "strength": 0.7,
  "scheduler": "UniPCMultistepScheduler",
  "lora_model": null,
  "tomesd": "yes",
  "use_karras_sigmas": "yes",
  "vae": null,
  "lora_strength": null,
  "embeddings_model": null,
  "seed": null,
  "webhook": null,
  "track_id": null
}

Request​

var myHeaders = new Headers();
myHeaders.append("Content-Type", "application/json");

var raw = JSON.stringify({
  "key": "",
  "model_id": "your_model_id",
  "prompt": "a cat sitting on a bench",
  "negative_prompt": null,
  "init_image": "https://raw.githubusercontent.com/CompVis/stable-diffusion/main/data/inpainting_examples/overture-creations-5sI6fQgYIuo.png",
  "mask_image": "https://raw.githubusercontent.com/CompVis/stable-diffusion/main/data/inpainting_examples/overture-creations-5sI6fQgYIuo_mask.png",
  "width": "512",
  "height": "512",
  "samples": "1",
  "steps": "21",
  "safety_checker": "no",
  "enhance_prompt": "yes",
  "guidance_scale": 7.5,
  "strength": 0.7,
  "scheduler": "PNDMScheduler",
  "lora_model": null,
  "tomesd": "yes",
  "use_karras_sigmas": "yes",
  "vae": null,
  "lora_strength": null,
  "embeddings_model": null,
  "seed": null,
  "webhook": null,
  "track_id": null
});

var requestOptions = {
  method: 'POST',
  headers: myHeaders,
  body: raw,
  redirect: 'follow'
};

fetch("https://modelslab.com/api/v6/images/inpaint", requestOptions)
  .then(response => response.text())
  .then(result => console.log(result))
  .catch(error => console.log('error', error));

Response​

Example Response
{
  "status": "success",
  "generationTime": 20.970642805099487,
  "id": 13446970,
  "output": [
    "https://pub-8b49af329fae499aa563997f5d4068a4.r2.dev/generations/dc639bd6-d605-42c7-950e-48c531124d0d-0.png"
  ],
  "meta": {
    "prompt": " a cat sitting on a bench DSLR photography, sharp focus, Unreal Engine 5, Octane Render, Redshift, ((cinematic lighting)), f/1.4, ISO 200, 1/160s, 8K, RAW, unedited, symmetrical balance, in-frame",
    "model_id": "midjourney-v4-painta",
    "scheduler": "PNDMScheduler",
    "safetychecker": "no",
    "negative_prompt": " ((out of frame)), ((extra fingers)), mutated hands, ((poorly drawn hands)), ((poorly drawn face)), (((mutation))), (((deformed))), (((tiling))), ((naked)), ((tile)), ((fleshpile)), ((ugly)), (((abstract))), blurry, ((bad anatomy)), ((bad proportions)), ((extra limbs)), cloned face, glitchy, ((extra breasts)), ((double torso)), ((extra arms)), ((extra hands)), ((mangled fingers)), ((missing breasts)), (missing lips), ((ugly face)), ((fat)), ((extra legs))",
    "W": 512,
    "H": 512,
    "guidance_scale": 7.5,
    "init_image": "https://raw.githubusercontent.com/CompVis/stable-diffusion/main/data/inpainting_examples/overture-creations-5sI6fQgYIuo.png",
    "mask_image": "https://raw.githubusercontent.com/CompVis/stable-diffusion/main/data/inpainting_examples/overture-creations-5sI6fQgYIuo_mask.png",
    "steps": 21,
    "n_samples": 1,
    "full_url": "no",
    "upscale": "no",
    "seed": 1343687916,
    "outdir": "out",
    "file_prefix": "dc639bd6-d605-42c7-950e-48c531124d0d"
  }
}