Publishing & rights
Connect files to records.
Match versions of a work across partners and associate identifiers with rights and descriptive metadata held in your own systems.
Learn / An introduction
The International Standard Content Code (ISCC) is a similarity-aware code. It is calculated from a media file, giving people and systems a common way to identify and compare content.
01 / A different kind of content identity
An ISBN refers to a publication. A DOI points to a registered object. An ISCC starts with the digital file itself.
Run the same algorithms on the same input and you can independently reproduce its code. You do not need to apply for an identifier or ask any organization to issue one.
ISCC combines similarity-preserving hashes with a cryptographic checksum. That lets you look for related content and distinguish an exact copy. It complements identifiers such as ISBN, DOI and URN by connecting records to the content they describe.
02 / Inside the code
Each component examines a different aspect of a file. Select a layer to see what it tells you.
Descriptive similarity
Compares descriptive metadata, such as a title and description. Similar descriptions can produce similar Meta-Codes.
Semantic similarity
Compares the meaning of content rather than its form. A paraphrase, a translation or a different picture of the same subject can produce similar Semantic-Codes.
Semantic-Code is implemented for text and images but not yet standardized.
Perceptual similarity
Compares the perceptual content of text, images, audio or video. It can reveal similarities after changes such as resizing or re-encoding, depending on the content and the transformation.
Binary similarity
Compares the binary data of files, whatever their media type. Shared sequences of bytes can remain recognizable when other parts of a file change.
Exact file identity
Uses a cryptographic checksum to identify the precise sequence of bytes. A change to the file changes its Instance-Code. This layer supports exact matching and integrity checks.
A modular standard. Data-Code and Instance-Code form the minimum ISCC-CODE. Metadata and content components add other perspectives. A Semantic-Code type is reserved in ISO 24138:2024; its algorithms are not standardized in that edition.
03 / In practice
Publishing & rights
Match versions of a work across partners and associate identifiers with rights and descriptive metadata held in your own systems.
Libraries & archives
Cluster similar digital objects, find duplicates and connect collections whose local catalogue identifiers differ.
Media & software
Support version management, duplicate detection and integrity checking in digital asset workflows.
A useful distinction
An ISCC does not establish authorship, ownership or authenticity. Those claims need evidence and context beyond the code itself.
04 / Common questions
The complete code can change. Similarity-preserving ISCC-UNITs may remain the same or close after certain edits, while the Instance-Code changes with the bytes. Matching depends on the media type, transformation and comparison threshold.
No. Generating and comparing ISCC-CODEs does not require either. Systems can use ISCCs locally or exchange them with other organizations.
The ISCC Discovery Protocol (IDP) is a new draft specification for declaring and discovering signed ISCCs to find associated metadata across existing authoritative registries.
Read the ISCC Enhancement Proposals ↗Yes. Open source tools can run in your own environment. Use the reference implementation for the standardized algorithms, or the SDK to work with media files.
Find developer resources →The standard is published as ISO 24138:2024. Its open source reference implementation is available as the standard's Electronic Insert, with documentation and code online.
Explore the standard →Put it into practice
Generate its code. Try a different version. Discover what connects.