# Token Economics

> The cost structure of agent interactions, where every token of input and output is billed — making concise, structured content a direct cost saving.

_The Agentic Web Lexicon · /glossary/token-economics · [JSON](/api/glossary/token-economics) · [all The Agentic Web Lexicon](/glossary)_

- **term:** Token Economics
- **category:** knowledge-memory
- **short_def:** The cost structure of agent interactions, where every token of input and output is billed — making concise, structured content a direct cost saving.
- **long_def:** Because agents pay per token, a markdown twin that is ~90% smaller than its HTML equivalent is not just faster but cheaper to consume. Agent-friendly design is partly an economic argument.
- **see_also:** markdown-twin, content-negotiation
- **etymology_origin:** — (verify-against-primary-at-build)
- **related_to:** markdown-twin, content-negotiation, rag
- **contrast_with:** Unlike blockchain 'tokenomics' (the supply and incentive design of a crypto token), token economics here means the per-token billing of LLM input and output — a content-cost argument, not a crypto one.
- **example:** A markdown twin is roughly 90% smaller than its HTML equivalent in tokens, so serving it directly lowers the per-token cost an agent pays to read the page.
- **source:** https://en.wikipedia.org/wiki/Large_language_model
- **status:** active
- **why_it_matters:** Token economics turns agent-readiness into a cost argument: leaner, structured content is literally cheaper for an agent to consume, which influences whether agents prefer your site.
- **sameAs:** —
- **bridge_entity:** agent-readiness
- **last_verified:** 2026-06-15
- **md_twin:** /glossary/token-economics.md