Universal Metadata Extraction

Media Workflows

Automatic metadata extraction for every image and video format. EXIF, XMP, IPTC, AI workflows—all captured without manual tagging. Privacy-aware export keeps you compliant.

  • 8 Formats Supported
  • <5s Extraction Time
  • Privacy-Aware Export

Supported Formats

Full metadata extraction, stripping, and injection for all major formats

Image Formats

FormatExtractStripInject
PNG
JPEG
WebP
TIFF

Extracted: EXIF, XMP, IPTC, ComfyUI workflows, Midjourney prompts

Injected: IPTC 2025.1 AI fields, custody chain

Video Formats

FormatExtractStripInject
MP4
MOV
WebM
MKV

Extracted: Duration, codec, framerate, bitrate, ComfyUI workflows

Method: MP4 atom parsing, Matroska EBML parsing

What We Extract

Comprehensive metadata capture from every file

AI Workflows

ComfyUI workflow graphs, Midjourney prompts, Stable Diffusion parameters, seeds, CFG scale, steps

Standard Metadata

EXIF (camera, GPS, timestamps), XMP (Adobe, rights), IPTC (creator, keywords, descriptions)

Video Technical

Duration, dimensions, framerate, bitrate, codec (H.264, H.265, VP9), encoder information

C2PA Detection

Content Authenticity Initiative manifest detection for provenance verification

Visual Features

Dimensions, aspect ratio, color profile, bit depth, resolution

Semantic Embeddings

Vector representations for natural language search across your entire asset library

Privacy-Aware Export

Share your work without exposing sensitive details. Our export pipeline strips what you want hidden and injects compliance metadata for EU AI Act and California SB 942.

Strippable (Category A)

  • Prompts, seeds, CFG scale
  • Model names, LoRA configs
  • GPS coordinates
  • Personal information
  • Full workflow JSON

Injected (Category C)

  • AISystemUsed indicator
  • DigitalSourceType
  • AIPromptInformation (generic)
  • Creator attribution
  • Custody chain event
Share PresetPortfolio PresetClient PresetArchive PresetCustom

Learn more about AI Transparency & Compliance →

What You Get

Fast Extraction

<5s for images, <10s for video. Batch processing at 500 files/minute.

Searchable Library

Full-text search across all extracted metadata. Find any asset instantly.

Cloud Storage

Free storage for your media files with all metadata preserved.

Lineage Tracking

Track every version, variation, and remix. Full provenance history.

Auto-Tagging

AI-generated tags from metadata. Zero manual entry required.

Desktop Sync

Monitor folders automatically. New files uploaded with full extraction.

Frequently Asked Questions

What metadata standards does Numonic support?

Numonic extracts and works with all major metadata standards: EXIF (camera settings, GPS, timestamps), XMP (Adobe metadata, rights management), IPTC (photo metadata, creator info), and the new IPTC 2025.1 AI disclosure fields. For video, we parse MP4 atoms and Matroska EBML containers.

How fast is metadata extraction?

Image extraction completes in under 5 seconds per asset. Video extraction takes under 10 seconds. For batch imports, we process approximately 500 files per minute in streaming mode. Success rate is over 95% for well-formed files.

Does stripping metadata degrade quality?

No. Our metadata stripping is lossless and operates at the container level only. We remove metadata chunks/atoms without re-encoding the media data. Your image and video quality remains exactly the same—only the metadata is modified.

What is IPTC 2025.1 injection?

IPTC 2025.1 is the latest photo metadata standard with specific fields for AI-generated content disclosure. Numonic injects fields like AISystemUsed, DigitalSourceType, and AIPromptInformation to help you comply with EU AI Act Article 50 and California SB 942 requirements. Learn more on our AI Transparency page.

Universal Metadata Extraction

Stop tagging manually. Let Numonic extract everything automatically.

Explore More