How does this tool detect AI-generated images?

We use a dual-engine analysis: Engine 1 checks the image's binary data for C2PA cryptographic credentials or Stable Diffusion/Midjourney parameters. Engine 2 analyzes the pixel data locally, examining high-frequency noise distribution, upsampling grid artifacts, and edge smoothness to evaluate probability even when metadata is stripped.

Can it detect images forwarded on WeChat, Weibo, or other social media?

Social platforms compress images and strip metadata. While the metadata engine will fail, our pixel feature engine will still scan for upsampling textures and high-frequency noise anomalies. Though compression reduces accuracy, it can still provide a valuable likelihood estimation.

Are my images uploaded to any servers?

Never. This tool runs 100% in your browser. All decoding, metadata parsing, and pixel analyses are done on your local device. Your files never leave your computer.

HomeBlog

AI Image Detector & Analyzer

Deeply detect if an image is generated by AI and extract parameters/workflows using the dual-engine of 'Metadata Credentials + Pixel Feature Analysis'.

Dual-Engine Detection 100% Local Privacy Parameter & Workflow Extractor

Drag & drop images here, or click to select

Supports PNG, JPG, WebP — processed entirely in your browser

Verdict

Waiting for Detection

Please upload or drag an image on the left. The system will automatically perform dual-engine identification and parameter extraction.

Pending

Pixel Feature Engine Report

Analysis of pixel-level noise distribution, upsampling periodic grids, and edge smoothness metrics will be displayed here.

Extracted AI Metadata

If the image contains AI metadata (like prompts, model, seed, and workflows), the parsed parameters will be restored and shown here.

How to Use AI Image Detector

Three simple steps.

Upload Image

Drag & drop or click to upload PNG, JPG, or WebP images.

Dual-Engine Analysis

The tool reads binary metadata while scanning pixels for noise anomalies and periodic checkerboard patterns.

View Comprehensive Report

Get AI probability estimates, software attribution, prompts, and ComfyUI workflow JSON.

Core Technology & Security

Deeply detect if an image is generated by AI and extract parameters/workflows using the dual-engine of 'Metadata Credentials + Pixel Feature Analysis'.

Dual-Engine Detection

Combines encrypted metadata verification with pixel-level high-frequency noise and upsampling artifact scanning.

100% Local Privacy

All image reading and pixel analysis are performed locally via JavaScript in your browser. No image data is uploaded.

Parameter & Workflow Extractor

Extracts Stable Diffusion, ComfyUI, and NovelAI prompts, parameters, and workflows with one-click copying.

Frequently Asked Questions

Find answers to the most common questions.