CBOR

CBOR (Concise Binary Object Representation), standardized as RFC 8949, is a compact binary serialization format built on the JSON data model. It represents the same maps, arrays, strings, and numbers as JSON, but encodes them as bytes rather than text, and it adds types JSON lacks—such as binary strings and a tagging mechanism for richer semantics. Its explicit design goals are small code size, small message size, and extensibility without version negotiation.

• JSON-compatible data model - Anything expressible in JSON maps cleanly to CBOR, easing conversion between the two.
• Compact binary encoding - Values are packed as bytes, cutting payload size and parsing cost versus text.
• Built for constrained environments - Minimal parser footprint makes it a fit for IoT devices and low-bandwidth links.
• Tags for extended types - A tag registry layers semantics like dates, big numbers, and URIs onto the core types.

CBOR shows up wherever JSON’s readability is not worth its size on the wire—embedded systems, IoT telemetry, and security protocols like COSE and WebAuthn that need deterministic, compact encoding. For API designers it is the pragmatic middle ground between human-friendly JSON and heavier binary schemes: the same shapes developers already model, delivered as bytes. As more machine-to-machine and agentic traffic flows between constrained endpoints, that combination of familiar model and small footprint keeps CBOR relevant.