DeMCP
DMCPMạng MCP phi tập trung
Buổi biểu diễn
Chỉ số Phổ Biến RD
Chỉ số Tăng Trưởng RD
So sánh
Theo dõi cập nhật
Theo dõi danh sách
Nhân vật
Về DeMCP
DeMCP is a decentralised Model Context Protocol network positioned as infrastructure between artificial intelligence and Web3. It addresses three problems in that ecosystem: most services still rely on standard input and output for communication, which makes distributed deployment and elastic scaling hard; remote services lack credible discovery and auditing; and the people who build services struggle to be paid according to usage.
In practice the network first converts standard input and output communication into remotely callable form through a server-sent events proxy layer and builds a service registry with health checks and basic scoring; services then run inside trusted execution environments with remote attestation and encrypted transport, while service identity, verification and reputation are written to a blockchain registry. Billing and incentives run on a token, with developers taking a share based on how often their services are called, and users able to pay in USDT and USDC and reach several major language models through one integration. Governance lets token holders vote on protocol upgrades, fee changes and new chain integrations.
That month the DeMCP team said more than 50 verified MCP services were integrated, spanning on-chain data, DeFi, social automation and LLM gateways, with over 10 hosted model endpoints and payment in USDT and USDC.
Architecturally, DeMCP spawns a stateless server instance for each API call, uses TEE remote attestation to prove model execution was not tampered with, and records provider identity and reputation in an on-chain registry. Revenue splits 70% to service developers, 15% to node operators, 10% to the ecosystem fund and 5% to burns. From January 2026 several AI and Web3 infrastructure studies cited it as a leading decentralized MCP project, then valued at roughly $1.6 million.
DeMCP will pursue a dedicated MCP chain and deeper TEE verification and registry integration, build a token-based reputation and review system, and lower the entry barrier through SDKs, a simulation sandbox and an LLM subsidy program.