opencomic-ai-bin

v1.5.0

Published

5 days ago

Upscale and dehalftone images using AI models like Real-CUGAN, Real-ESRGAN, Waifu2x and Upscayl.

0High
0Medium
0Low

ollm

opencomic upscaling halftone dehalftone real-cugan real-esrgan waifu2x upscayl

opencomic-ai-bin

This package provides pre-built binaries (realcugan, waifu2x, and upscayl) through a simple Node.js API.

Model files are not included, they are downloaded automatically when needed.

If you prefer to install all models upfront, use the opencomic-ai-models package.

Installation

npm install opencomic-ai-bin

Methods

This package can be used using CommonJS or ES Module.

// CommonJS
const OpenComicAI = require('opencomic-ai-bin');

// ES Module
import OpenComicAI from 'opencomic-ai-bin';

Simple example of package usage

import OpenComicAI from 'opencomic-ai-bin';
import sharp from 'sharp'; // This is optional, only needed if you want to keep ICC profile or use a YOLO26 model

(async () => {

	// Set the base directory for binary paths, for example change to app.asar.unpacked path in Electron apps
	OpenComicAI.setDirname(OpenComicAI.__dirname.replace(/app(-(?:arm64|x64))?\.asar/, 'app$1.asar.unpacked'));

	// Models path, if the model is not found in this folder, it will be downloaded
	OpenComicAI.setModelsPath('./models');

	// Set sharp instance
	OpenComicAI.setSharp(sharp); 

	// Keep ICC profile from input image, requires sharp instance
	OpenComicAI.keepIccProfile('rgb16');

	await OpenComicAI.pipeline('./input.jpg', './output.jpg', [
		{
			model: 'opencomic-ai-descreen-hard-fast-v3',
		},
		{
			model: 'opencomic-ai-upscale-fast-v3',
			scale: 4,
		}
	], (progress) => {

		console.log(`Processing: ${Math.round(progress * 100)}%`);

	}, {
		start: () => {

			console.log(`Start download`);

		},
		progress: (progress) => {

			console.log(`Downloading: ${Math.round(progress * 100)}%`);

		},
		end: () => {

			console.log(`End download`);

		},
	});

})();

OpenComicAI.setDirname

Set the base directory for binary paths.

OpenComicAI.setDirname(dirname: string): void

OpenComicAI.setModelsPath

Set the directory where models will be downloaded and stored.

OpenComicAI.setModelsPath(path: string): void

OpenComicAI.setCustomModel

OpenComicAI.setCustomModel(model: ModelObject): void

The model argument follows the ModelObject interface. The key, name, upscaler, scales, and files properties are required when registering a custom model.

The model files must already exist locally. Their paths are resolved as follows:

<modelsPath>/<type>/<folder>/<file>

For example, with type: 'custom' and folder: './models', place the model files in <modelsPath>/custom/models/. The .bin and .param files must use the exact names listed in files.

const customModel = 'my-custom-model-1x';

// Use a path appropriate for your installation. The selected directory must
// resolve to the folder containing the platform binary directories.
OpenComicAI.setDirname('/path/to/opencomic-ai-bin/runtime');
OpenComicAI.setModelsPath('/path/to/models');

OpenComicAI.setCustomModel({
	key: customModel,
	name: 'My Custom Model',
	type: 'custom',
	upscaler: 'upscayl',
	scales: [1],
	noise: undefined,
	latency: 1,
	folder: './models',
	files: [
		`${customModel}.bin`,
		`${customModel}.param`,
	],
});

await OpenComicAI.image('./input.png', './output.png', {
	model: customModel,
});

setDirname() is only needed when the custom model uses a binary outside the package's default runtime location. The upscayl binary must be available for the current platform, and the model name passed to image() or pipeline() must match the key used during registration.

OpenComicAI.setConcurrentDaemons

Set the maximum number of concurrent daemons (upscayl models only), 0 disables daemons.

Daemons help speed up processing loading model only once and not for each image. See comparative performance.

OpenComicAI.setConcurrentDaemons(count: number = 3): void

OpenComicAI.setDaemonIdleTimeout

Set the idle timeout for daemons in milliseconds (upscayl models only).

OpenComicAI.setDaemonIdleTimeout(timeout: number = 60000): void

OpenComicAI.closeAllDaemons

Close all running daemons (upscayl models only).

OpenComicAI.closeAllDaemons(): void

OpenComicAI.setSharp

Set a sharp instance.

OpenComicAI.setSharp(sharp: any): void

OpenComicAI.keepIccProfile

Keep the ICC profile from the input image in the output image. This requires a sharp instance configured with setSharp().

The output is converted to the specified pipeline colourspace before the input ICC profile is applied.

OpenComicAI.keepIccProfile(pipelineColourspace: string = 'rgb16'): void

OpenComicAI.__dirname

Get the base directory for binary paths.

OpenComicAI.__dirname: string

OpenComicAI.models

Object containing all available models organized by type (upscale, descreen, descreen-mask, artifact-removal, panels, and custom).

OpenComicAI.models: Record<ModelType, Record<string, ModelObject>>

OpenComicAI.modelsList

Array of all available model keys.

OpenComicAI.modelsList: Model[]

OpenComicAI.modelsTypeList

Models organized by type.

OpenComicAI.modelsTypeList: Record<ModelType, Model[]>

OpenComicAI.modelsPath

Current path where models are stored.

OpenComicAI.modelsPath: string | undefined

OpenComicAI.binary

Get the path to the binary executable for a model.

OpenComicAI.binary(model: Model): string

OpenComicAI.model

Get detailed information about a specific model.

OpenComicAI.model(model: Model): ModelObject

OpenComicAI.preload

Preload the model to daemon (upscayl models only) or download the model if not available locally.

OpenComicAI.preload(steps: OpenComicAIOptions[], downloading?: Downloading): Promise<void>

OpenComicAI.image

Process an image using a single AI model. If no model is specified, the default model is used.

The destination directory is created automatically when it does not exist.

OpenComicAI.image(
	source: string,
	dest: string,
	options?: OpenComicAIOptions,
	progress?: ((progress: number) => void) | false
): Promise<string>

OpenComicAI.pipeline

Process an image through one or more AI models.

OpenComicAI.pipeline(
	source: string,
	dest: string,
	steps: OpenComicAIOptions[],
	progress?: (progress: number) => void,
	downloading?: Downloading
): Promise<string>

OpenComicAI.closest

Returns the value in the array closest to the target value.

OpenComicAI.closest(array: number[], target: number): number

Types

`Model`

type Model =
	| 'opencomic-ai-upscale-fast-v3'
	| 'opencomic-ai-upscale-balanced-v3'
	...

Model aliases

The following aliases identify models for specific processing stages:

type ModelUpscale = keyof typeof OpenComicAI.models.upscale;
type ModelArtifactRemoval = keyof typeof OpenComicAI.models['artifact-removal'];
type ModelMask = keyof typeof OpenComicAI.models['descreen-mask'];
type ModelPanels = keyof typeof OpenComicAI.models.panels;
type ModelCustom = string;

`ModelType`

type ModelType =
	| 'upscale'
	| 'descreen'
	| 'descreen-mask'
	| 'artifact-removal'
	| 'panels'
	| 'custom';

`Upscaler`

type Upscaler = 'realcugan' | 'upscayl' | 'waifu2x';

`Speed`

type Speed = 'Very Fast' | 'Fast' | 'Medium' | 'Slow' | 'Very Slow';

`OpenComicAIOptions`

interface OpenComicAIOptions {
	model?: Model;
	noise?: 0 | 1 | 2 | 3;
	scale?: number;
	tileSize?: number | 'auto';
	maxTileSize?: number;
	memorySafePercentage?: number;
	gpuId?: string;
	threads?: number;
	tta?: boolean;
	fp32?: boolean;
	keepBigHalftone?: OpenComicAIKeepBigHalftone;
}

`OpenComicAIKeepBigHalftone`

Options for preserving large halftone components while descreening an image.

interface OpenComicAIKeepBigHalftone {
	model: ModelMask | 'auto';
	minSize?: number;
	minPixels?: number;
	artifactRemoval?: OpenComicAIArtifactRemoval;
}

minSize is the minimum component size as a percentage of the image height. minPixels filters out components smaller than the specified number of pixels. artifactRemoval can be used to process the selected regions with an artifact-removal model.

`OpenComicAIArtifactRemoval`

interface OpenComicAIArtifactRemoval {
	model: ModelArtifactRemoval | 'auto';
}

`OpenComicAIPanels`

Options for panel detection.

interface OpenComicAIPanels {
	model: ModelPanels | 'auto';
	minPixels?: number;
	keepAspectRatio?: boolean;
	upscale?: OpenComicAIUpscale;
}

upscale can be used to upscale the panel mask before detecting its components.

`ModelObject`

interface ModelObject {
	key?: Model;
	name: string;
	upscaler: Upscaler;
	type?: ModelType;
	tileSize?: number; // Force tile size for this model, overrides auto tile estimation
	tileSizeFromMem128?: number; // Model memory usage in MB measured at tile=128 for auto tile estimation
	scales: number[];
	noise: number[] | undefined;
	latency: number; // From 0.5 (Fastest model) to 10 (Slowest model)
	speed?: Speed;
	folder: string;
	path?: string;
	files: string[];
	scaleFiles?: Record<number, Model>;
	supportCurrentPlatform?: boolean;
}

`Downloading`

interface Downloading {
	start?: () => void;
	progress?: (progress: number) => void;
	end?: () => void;
}

Models Info

Artifact removal

Model | Name | Upscaler | Source ------|------|----------|------- opencomic-ai-artifact-removal-balanced-v3-ps | OpenComic AI Artifact Removal Balanced v3 | upscayl | ollm/opencomic-ai-training opencomic-ai-artifact-removal-quality-v3-ps | OpenComic AI Artifact Removal Quality v3 | upscayl | ollm/opencomic-ai-training opencomic-ai-artifact-removal-compact-v2 | OpenComic AI Artifact Removal Fast v2 | upscayl | ollm/opencomic-ai-training opencomic-ai-artifact-removal-lite-v2 | OpenComic AI Artifact Removal Balanced v2 | upscayl | ollm/opencomic-ai-training opencomic-ai-artifact-removal-v2 | OpenComic AI Artifact Removal Quality v2 | upscayl | ollm/opencomic-ai-training opencomic-ai-artifact-removal-compact | OpenComic AI Artifact Removal Fast | upscayl | ollm/opencomic-ai-training opencomic-ai-artifact-removal-lite | OpenComic AI Artifact Removal Balanced | upscayl | ollm/opencomic-ai-training opencomic-ai-artifact-removal | OpenComic AI Artifact Removal Quality | upscayl | ollm/opencomic-ai-training 1x_NMKD-Jaywreck3-Lite_320k | NMKD Jaywreck3 Lite | upscayl | NMKD.de 1x_NMKD-Jaywreck3-Soft-Lite_320k | NMKD Jaywreck3 Soft Lite | upscayl | NMKD.de 1x-SaiyaJin-DeJpeg | SaiyaJin DeJpeg | upscayl | OpenModelDB 1x_JPEGDestroyerV2_96000G | JPEG Destroyer V2 | upscayl | Hugging Face

Descreen

Model | Name | Upscaler | Source ------|------|----------|------- opencomic-ai-descreen-hard-fast-v3-256 | OpenComic AI Descreen Hard Fast v3 | upscayl | ollm/opencomic-ai-training opencomic-ai-descreen-hard-balanced-v3-256 | OpenComic AI Descreen Hard Balanced v3 | upscayl | ollm/opencomic-ai-training opencomic-ai-descreen-hard-compact-v2 | OpenComic AI Descreen Hard Fast v2 | upscayl | ollm/opencomic-ai-training opencomic-ai-descreen-hard-lite-v2 | OpenComic AI Descreen Hard Balanced v2 | upscayl | ollm/opencomic-ai-training opencomic-ai-descreen-hard-v2 | OpenComic AI Descreen Hard Quality v2 | upscayl | ollm/opencomic-ai-training opencomic-ai-descreen-hard-compact | OpenComic AI Descreen Hard Fast | upscayl | ollm/opencomic-ai-training opencomic-ai-descreen-hard-lite | OpenComic AI Descreen Hard Balanced | upscayl | ollm/opencomic-ai-training opencomic-ai-descreen-hard | OpenComic AI Descreen Hard Quality | upscayl | ollm/opencomic-ai-training 1x_halftone_patch_060000_G | Halftone Patch 060000 G | upscayl | NMKD.de 1x_wtp_descreenton_compact | WTP DescreenTon Compact | upscayl | OpenModelDB

Upscale

Model | Name | Upscaler | Source ------|------|----------|------- opencomic-ai-upscale-fast-v3 | OpenComic AI Upscale Fast v3 | upscayl | ollm/opencomic-ai-training opencomic-ai-upscale-balanced-v3 | OpenComic AI Upscale Balanced v3 | upscayl | ollm/opencomic-ai-training opencomic-ai-upscale-compact-v2 | OpenComic AI Upscale Fast v2 | upscayl | ollm/opencomic-ai-training opencomic-ai-upscale-lite-v2 | OpenComic AI Upscale Balanced v2 | upscayl | ollm/opencomic-ai-training opencomic-ai-upscale-v2 | OpenComic AI Upscale Quality v2 | upscayl | ollm/opencomic-ai-training opencomic-ai-upscale-compact | OpenComic AI Upscale Fast | upscayl | ollm/opencomic-ai-training opencomic-ai-upscale-lite | OpenComic AI Upscale Balanced | upscayl | ollm/opencomic-ai-training opencomic-ai-upscale | OpenComic AI Upscale Quality | upscayl | ollm/opencomic-ai-training realcugan | RealCUGAN | realcugan | Moebytes/waifu2x realesr-animevideov3 | RealESR AnimeVideo v3 | upscayl | xinntao/Real-ESRGAN realesrgan-x4plus | RealESRGAN x4 Plus | upscayl | xinntao/Real-ESRGAN realesrgan-x4plus-anime | RealESRGAN x4 Plus Anime | upscayl | xinntao/Real-ESRGAN realesrnet-x4plus | RealESRNet x4 Plus | upscayl | xinntao/Real-ESRGAN waifu2x-models-cunet | Waifu2x CUnet | waifu2x | Moebytes/waifu2x waifu2x-models-upconv | Waifu2x UpConv | waifu2x | Moebytes/waifu2x 4x-WTP-ColorDS | WTP ColorDS | upscayl | OpenModelDB remacri-4x | Remacri | upscayl | upscayl/upscayl ultramix-balanced-4x | Ultramix Balanced | upscayl | upscayl/upscayl ultrasharp-4x | Ultrasharp | upscayl | upscayl/upscayl 4xInt-RemAnime | Int-RemAnime | upscayl | Phhofm/models AI-Forever_x4plus | AI-Forever x4plus | upscayl | Hugging Face 4xNomosWebPhoto_esrgan | Nomos Web Photo ESRGAN | upscayl | OpenModelDB 4xHFA2k | HFA2k | upscayl | upscayl/custom-models 4xLSDIRCompactC3 | LSDIR Compact C3 | upscayl | upscayl/custom-models 4xLSDIRplusC | LSDIR Plus C | upscayl | upscayl/custom-models 4x_NMKD-Siax_200k | NMKD Siax | upscayl | upscayl/custom-models 4xNomos8kSC | Nomos 8k SC | upscayl | upscayl/custom-models RealESRGAN_General_WDN_x4_v3 | RealESRGAN General WDN v3 | upscayl | upscayl/custom-models RealESRGAN_General_x4_v3 | RealESRGAN General v3 | upscayl | upscayl/custom-models uniscale_restore_x4 | Uniscale Restore x4 | upscayl | upscayl/custom-models unknown-2.0.1 | Unknown 2.0.1 | upscayl | upscayl/custom-models

Daemon comparative performance

Table (10 images 512x512px)

Model | Disabled | Enabled | Disabled vs Enabled -------------------------------|-----------------------|----------------------|-------- OpenComic AI Upscale Lite | 52.087s | 7.646s | 6.81x RealESRGAN x4 Plus | 73.273s | 23.199s | 3.16x

OpenComic AI Upscale Lite

Disabled

Processing image 1/10 for model: OpenComic AI Upscale Lite: 2.931s
Processing image 2/10 for model: OpenComic AI Upscale Lite: 4.464s
Processing image 3/10 for model: OpenComic AI Upscale Lite: 5.594s
Processing image 4/10 for model: OpenComic AI Upscale Lite: 5.561s
Processing image 5/10 for model: OpenComic AI Upscale Lite: 5.521s
Processing image 6/10 for model: OpenComic AI Upscale Lite: 5.531s
Processing image 7/10 for model: OpenComic AI Upscale Lite: 5.534s
Processing image 8/10 for model: OpenComic AI Upscale Lite: 5.555s
Processing image 9/10 for model: OpenComic AI Upscale Lite: 5.496s
Processing image 10/10 for model: OpenComic AI Upscale Lite: 5.898s
Model: OpenComic AI Upscale Lite, Latency: 52.087s

Enabled

Preloading model... OpenComic AI Upscale Lite
Preload model: OpenComic AI Upscale Lite: 473.897ms
Processing image 1/10 for model: OpenComic AI Upscale Lite: 725.474ms
Processing image 2/10 for model: OpenComic AI Upscale Lite: 716.566ms
Processing image 3/10 for model: OpenComic AI Upscale Lite: 716.221ms
Processing image 4/10 for model: OpenComic AI Upscale Lite: 715.214ms
Processing image 5/10 for model: OpenComic AI Upscale Lite: 717.894ms
Processing image 6/10 for model: OpenComic AI Upscale Lite: 715.236ms
Processing image 7/10 for model: OpenComic AI Upscale Lite: 716.296ms
Processing image 8/10 for model: OpenComic AI Upscale Lite: 714.025ms
Processing image 9/10 for model: OpenComic AI Upscale Lite: 718.653ms
Processing image 10/10 for model: OpenComic AI Upscale Lite: 714.792ms
Model: OpenComic AI Upscale Lite, Latency: 7.646s

RealESRGAN x4 Plus

Disabled

Processing image 1/10 for model: RealESRGAN x4 Plus: 6.473s
Processing image 2/10 for model: RealESRGAN x4 Plus: 7.809s
Processing image 3/10 for model: RealESRGAN x4 Plus: 7.690s
Processing image 4/10 for model: RealESRGAN x4 Plus: 7.470s
Processing image 5/10 for model: RealESRGAN x4 Plus: 6.620s
Processing image 6/10 for model: RealESRGAN x4 Plus: 7.390s
Processing image 7/10 for model: RealESRGAN x4 Plus: 7.423s
Processing image 8/10 for model: RealESRGAN x4 Plus: 7.541s
Processing image 9/10 for model: RealESRGAN x4 Plus: 7.536s
Processing image 10/10 for model: RealESRGAN x4 Plus: 7.321s
Model: RealESRGAN x4 Plus, Latency: 73.273s

Enabled

Preloading model... RealESRGAN x4 Plus
Preload model: RealESRGAN x4 Plus: 1.165s
Processing image 1/10 for model: RealESRGAN x4 Plus: 2.217s
Processing image 2/10 for model: RealESRGAN x4 Plus: 2.196s
Processing image 3/10 for model: RealESRGAN x4 Plus: 2.203s
Processing image 4/10 for model: RealESRGAN x4 Plus: 2.191s
Processing image 5/10 for model: RealESRGAN x4 Plus: 2.200s
Processing image 6/10 for model: RealESRGAN x4 Plus: 2.203s
Processing image 7/10 for model: RealESRGAN x4 Plus: 2.216s
Processing image 8/10 for model: RealESRGAN x4 Plus: 2.222s
Processing image 9/10 for model: RealESRGAN x4 Plus: 2.188s
Processing image 10/10 for model: RealESRGAN x4 Plus: 2.194s
Model: RealESRGAN x4 Plus, Latency: 23.199s

Credits

This project uses the following AI models and upscalers:

Related Projects

Citation (opencomic-ai-training)

If you use this project, please cite it as follows.

@software{opencomic_ai,
    author = {Llopart Mora, Oleguer},
    title = {{OpenComic AI}},
    year = {2026},
    version = {2.0},
    doi = {10.5281/zenodo.20802370},
    url = {https://doi.org/10.5281/zenodo.20802370}
}

Release models citation. To see the differences between versions, you can check the CHANGELOG.md file.

OpenComic AI v3.0

@software{opencomic_ai_models_v3_0,
    author = {Llopart Mora, Oleguer},
    title = {{OpenComic AI v3.0 Models}},
    year = {2026},
    version = {3.0},
    doi = {10.5281/zenodo.23245460},
    url = {https://doi.org/10.5281/zenodo.23245460}
}

OpenComic AI v2.0

@software{opencomic_ai_models_v2_0,
    author = {Llopart Mora, Oleguer},
    title = {{OpenComic AI v2.0 Models}},
    year = {2026},
    version = {2.0},
    doi = {10.5281/zenodo.22967095},
    url = {https://doi.org/10.5281/zenodo.22967095}
}

OpenComic AI v1.0

@software{opencomic_ai_models_v1_0,
    author = {Llopart Mora, Oleguer},
    title = {{OpenComic AI v1.0 Models}},
    year = {2026},
    version = {1.0.1},
    doi = {10.5281/zenodo.20802371},
    url = {https://doi.org/10.5281/zenodo.20802371}
}

License

| Component | License | |---|---| | Source code | MIT | | OpenComic AI model weights | CC BY 4.0 |

The source code of this repository is licensed under the MIT License.

OpenComic AI model weights are licensed under the CC BY 4.0.

See LICENSE and LICENSE_MODELS for details.

Published

Vulnerabilities

Links

Maintainers

Keywords

Readme

opencomic-ai-bin

Installation

Methods

Simple example of package usage

OpenComicAI.setDirname

OpenComicAI.setModelsPath

OpenComicAI.setCustomModel

OpenComicAI.setConcurrentDaemons

OpenComicAI.setDaemonIdleTimeout

OpenComicAI.closeAllDaemons

OpenComicAI.setSharp

OpenComicAI.keepIccProfile

OpenComicAI.__dirname

OpenComicAI.models

OpenComicAI.modelsList

OpenComicAI.modelsTypeList

OpenComicAI.modelsPath

OpenComicAI.binary

OpenComicAI.model

OpenComicAI.preload

OpenComicAI.image

OpenComicAI.pipeline

OpenComicAI.closest

Types

Model

Model aliases

ModelType

Upscaler

Speed

OpenComicAIOptions

OpenComicAIKeepBigHalftone

OpenComicAIArtifactRemoval

OpenComicAIPanels

ModelObject

Downloading

Models Info

Artifact removal

Descreen

Upscale

Daemon comparative performance

Table (10 images 512x512px)

OpenComic AI Upscale Lite

Disabled

Enabled

RealESRGAN x4 Plus

Disabled

Enabled

Credits

Related Projects

Citation (opencomic-ai-training)

OpenComic AI v3.0

OpenComic AI v2.0

OpenComic AI v1.0

License

`Model`

`ModelType`

`Upscaler`

`Speed`

`OpenComicAIOptions`

`OpenComicAIKeepBigHalftone`

`OpenComicAIArtifactRemoval`

`OpenComicAIPanels`

`ModelObject`

`Downloading`