AI Metadata Cleaner

File metadata is hidden data embedded inside files that most people never see. An image taken with your phone can contain:

• GPS coordinates — exact location where the photo was taken
• Device info — your phone model, serial number, firmware
• Timestamps — exactly when you created or modified the file
• Identity info — your name from device settings

When you share files online, this metadata travels with them, potentially revealing your location, identity, and device to anyone who checks.

AI image generators embed generation parameters in different ways:

• Stable Diffusion / AUTOMATIC1111: Stores full prompt, negative prompt, seed, CFG scale, sampler in PNG tEXt chunks under the key "parameters"
• ComfyUI: Stores JSON workflow data in PNG chunks
• NovelAI: Stores source tag identifying the service
• DALL-E / Midjourney: Some versions embed Software or Comment EXIF fields

This tool scans all these locations and flags AI-generated content automatically.

Absolutely not. This tool runs 100% in your browser using JavaScript's File API, Canvas API, and binary parsing. Your files are read directly into browser memory and never transmitted anywhere. No internet connection is required after the page loads. Your data stays on your device at all times.

Full read + clean: JPEG, PNG, WEBP, GIF, BMP — images are redrawn on canvas to strip all EXIF

Read + clean: DOCX, XLSX, PPTX — Office files are repackaged with metadata XML removed

Read only (metadata display): PDF, MP3, MP4, WAV, OGG, SVG — metadata is shown but cleaning these formats in pure browser JS is limited

Text files: TXT, RTF, SVG — scanned for embedded metadata or suspicious patterns

EXIF (Exchangeable Image File Format) is a standard for storing metadata in image files. It was designed to help cameras record technical photo details. Common EXIF fields include: camera make/model, lens info, aperture, shutter speed, ISO, focal length, GPS location, date/time, and software used. Modern smartphones automatically embed GPS coordinates in every photo, which can reveal exactly where you live, work, or travel.

Yes. PDF metadata often includes: Author name (from the computer's user account), creating software (Word version, Adobe version), company name, creation and modification dates, and revision history. Legal documents, contracts, and business files shared as PDFs frequently expose the author's identity and organization through this hidden data.

Microsoft Office files (DOCX, XLSX, PPTX) store extensive metadata in docProps/core.xml and docProps/app.xml inside the file. This includes: creator name, last modified by (username), creation date, last modified date, company name, revision number, total editing time, and the application version used. This tool reads these XML files and can regenerate the Office file with all metadata removed.

The risk score is calculated based on the sensitivity of metadata found:

• High Risk 🔴 — GPS coordinates, device serial numbers, or AI generation prompts with personal content
• Medium Risk 🟠 — Author name, company info, creation timestamps, camera model
• Low Risk 🟡 — Technical metadata like resolution, color space, file format version
• Clean ✅ — No sensitive metadata detected