Deal cards to agents.
Stack decks for context.
The magic is in the cards.
Self-contained information unit (~250 tokens)
Interactive buttons and triggers
Branch points and choices
Open Protocol. Reference Resolver.
Card Network is an open protocol (cardnetwork.org) — anyone can implement it. We run the reference resolver (api.cardnetwork.dev) — easiest way to get started.
cardnetwork.dev
Landing, demo, dashboard, pricing
api.cardnetwork.dev
Resolver + Registry API
cardnetwork.org
Open protocol spec (CC BY 4.0)
The Content Problem
Current content systems force a choice between human readability and machine efficiency.
Documents Are Monolithic
Long documents get chunked into fragments, losing context and relationships. AI agents struggle with where one concept ends and another begins.
Links Are Untyped
Hyperlinks say "there's a connection" but not what kind. Is it a sequence? A reference? A dependency? Navigation becomes guesswork.
AI Context Is Wasted
LLMs have limited context windows. Feeding them 10,000 tokens of a document when they need 250 tokens of relevant info is inefficient.
The C.A.R.D. Framework
"Context isn't a card. Context is a DECK."
ontext
Layered, composable from multiple cards. Context isn't static; it's built dynamically from what's relevant.
tomic
Each card is self-contained with a predictable schema. Small enough for any context window, complete enough to stand alone.
etrievable
Spatial and semantic addressing. Find cards not just by keyword, but by their relationships and place in the network.
eck
The managed working set in context. Agents build decks to solve problems, resizing and reshuffling as they reason.
Universal Card Addressing
Every card lives in a deck. Every deck has an owner. Resolve any card from any deck, anywhere in the network.
"Draw from any deck. Your agent requests card:@[stripe.com/docs/quickstart] — the resolver finds the deck, deals the card."
How Card Network Solves This
Cards as atomic units. Typed edges as relationships. Compositions as navigable structures.
Atomic Cards
Each card is self-contained, ~250 tokens. Perfect for small context windows, easy to cache, works offline.
Typed Edges
Relationships have meaning. Stack for sequences, link for references, branch for conditionals, depends for prerequisites.
Smart Compositions
Build stacks for tutorials, networks for knowledge bases, trees for decisions. Navigate visually or traverse programmatically.
Built for the Agentic Web
Card Network bridges human navigation and AI consumption, optimized for edge computing.
AI-Native
~250 tokens per card fits in any LLM context window. Typed edges enable graph traversal without inference.
Mobile-First
Cards sized like phone screens. Swipe through decks, tap to navigate, pinch to zoom out to network view.
Edge-Ready
Works on Raspberry Pi, mobile devices, IoT. Minimal compute, aggressive caching, offline-capable.
Open Protocol
CC BY 4.0 licensed. JSON Schema defined. Build your own tools, integrate anywhere, no lock-in.
Draw what you need
Instead of 10,000 tokens of document, draw exactly the 750 tokens (3 cards) you need to answer the question.
Wild Cards
Branch points and choices built-in. Query by edge type. Find all dependencies, follow all sequences.
Build your Hand
Agents don't just "retrieve" context — they build a hand. Budget tokens precisely, know exactly how many cards fit.
Face-up vs Face-down
Public cards are face-up. Private, encrypted cards are face-down. The protocol manages both seamlessly.
Start Free. Scale When Ready.
Your API key is your wallet. Your wallet builds your hand.
Anonymous
No signup required
Claimed
Email signup required
Funded
For power users
"The house doesn't always win. Publish a deck — earn when others draw."
Ready to Build with Cards?
Explore the interactive demo, read the specification, or start building today.
# Keep this key to build your hand across requests